How to Build & Sell Voice AI Agents in 2026 (ULTIMATE Guide)

How to Build and Sell Voice AI Agents

Introduction to the Masterclass

• This video aims to provide a comprehensive guide on building and selling voice AI agents, covering technical, business, and foundational aspects with a real-world use case from a law firm.

• The course includes both inbound qualification appointment setting and outbound voice agent functionalities for collecting overdue invoices.

Instructor's Background

• The instructor has 8 years of experience in AI, including roles as a data scientist at Nike and involvement in machine learning projects. Two years ago, they focused entirely on voice AI.

• They have deployed over 30 voice agents across various industries and conducted more than 600 hours of live coaching sessions, helping over 1,000 individuals create their first voice agent.

Course Structure Overview

• A roadmap is provided that outlines the build stage (foundations, platforms) on one side and the sales side (packaging, pricing, closing clients) on the other. Each phase includes checklists for clarity.

• Viewers are encouraged to take screenshots of the roadmap for reference throughout the tutorial. The checklist covers everything from deciding what to build to delivering it effectively to clients.

Learning Approach

• Participants are advised to pause when necessary due to the volume of information presented; transitions between topics will be indicated by loading bars for optimal break points.

• Active participation is encouraged—viewers should replicate steps alongside the instructor for deeper learning retention through practical application rather than passive observation.

Building Skills with Voice AI Systems

• The focus is not just on creating demos but developing full-fledged voice AI systems that can be packaged and sold; templates will be available via Skool for efficiency but skills remain paramount.

• Understanding what constitutes a voice AI agent is crucial: these systems interact with people over the phone in real-time (e.g., booking appointments or sending reminders). Examples illustrate practical applications of this technology.

Market Potential & Industry Relevance

• Current advancements make now an ideal time for deploying voice agents due to improved tools that enhance instruction-following capabilities and natural-sounding voice generation at reduced costs compared to human receptionists.

• Voice AI addresses significant business challenges such as staffing shortages by providing 24/7 availability while ensuring brand consistency in customer interactions—a growing necessity as businesses increasingly adopt these technologies.

Voice AI Tools: The Future of Business Communication

The Rise of Voice AI in Business

• Voice AI tools are becoming increasingly common, marking a shift from early adopters to the early majority by 2026, which will represent a significant market segment.

• Businesses should capitalize on the readiness of voice technology, focusing on profitable industries and specific use cases for voice AI.

Understanding Voice Agents

• A voice agent can answer calls instead of a human, utilizing speech-to-text models to understand what leads say.

• The process involves reasoning over text with a large language model (LLM), generating responses via text-to-speech technology in real-time.

Key Components of Voice Agents

• Recommended high-quality voices come from 11 Labs; users can select different voices based on preference.

• GPT 4.1 is suggested as an effective LLM for understanding conversation nuances and following instructions accurately.

Inbound vs. Outbound Calls

• Inbound agents handle customer calls for various purposes like appointment booking or customer support, requiring careful management of user information.

• Outbound agents call customers for reminders or follow-ups, emphasizing the importance of tone and approach to avoid sounding salesy.

Use Case: Family Law Firm

• The project focuses on building inbound and outbound agents specifically for family law firms to manage client interactions effectively.

• Inbound agents qualify potential clients needing legal assistance while outbound agents address overdue invoices and payment processing securely.

Why Target Law Firms?

• Law firms have high-value clients with significant lifetime value, making them ideal candidates for enhanced service through voice AI solutions.

• Existing workflows in law firms allow for clear processes that can be optimized using voice technology, leading to better client retention and satisfaction.

Voice Agents and Their Applications

Industry Applications of Voice Agents

• Voice agents can enhance various industries, such as healthcare, where inbound receptionists often miss calls while managing other tasks.

• The concept of "speed to lead" is crucial; it refers to contacting potential clients immediately after they express interest online, maximizing the likelihood of conversion.

• Effective voice agents can answer questions and book appointments in real-time, which is particularly beneficial for businesses running ads or focusing on SEO.

• Database reactivation is another industry-agnostic use case; many businesses fail to follow up with leads promptly, resulting in lost opportunities.

• Re-engaging cold leads can be seen as "free money" for businesses if approached correctly.

Building Versatile Voice Agent Systems

• The development of swappable components allows customization of voice agents by choosing different voices, LLMs (Language Learning Models), and tools tailored to specific business needs.

• A recommended tech stack includes Retell AI for voice agents, n8n for automation integration with CRMs, GoHighLevel as a CRM platform, and Cal.com for booking software.

• These platforms can be interconnected so that bookings made through one system reflect across others seamlessly.

Inbound vs. Outbound Use Cases

• For inbound calls, the process begins with identifying whether the caller is an existing customer using n8n to check against GoHighLevel's database.

• If the user is known, dynamic variables personalize the interaction; if not recognized, the agent qualifies them before scheduling an appointment via Cal.com.

Outbound Call Scenarios

• An example of outbound use involves payment reminders triggered by overdue invoices over 30 days through GoHighLevel.

• The conversation starts with acknowledging previous issues resolved and addressing outstanding payments directly during the call.

Customizing Client Interactions

• Understanding client preferences is essential when scheduling follow-up calls; some may prefer multiple attempts in a day while others might want spaced-out reminders based on their brand image.

• Transitioning from builder mode to consultant mode helps tailor solutions that align with client expectations and improve satisfaction.

Setting Up Platforms

• The next steps involve signing up for Retell AI using a referral link provided in the video description.

Getting Started with Retell: Setting Up Your Agent

Account Creation and Initial Setup

• The process begins by signing up for an account using your email address, followed by entering a verification code received via email.

• Users can purchase new phone numbers or connect existing ones through SIP trunking; batch calling allows multiple calls to be made simultaneously.

• Call history is crucial for analyzing past interactions to improve agent performance, while chat history provides insights into chatbot interactions.

Billing and Settings Overview

• Users are encouraged to set up billing information, including credit card details, although a free trial with $10 credits is available.

• There are options to import agents from files created in the session or manually create them for skill-building purposes.

Creating an Agent

• Manual creation of agents is recommended initially to develop skills rather than relying solely on templates; downloadable agents will be provided post-video.

• Key elements during agent setup include naming the agent, noting the agent ID, and understanding cost per minute (approximately 11 cents).

Configuration Options

• The current voice agent uses 11 Labs technology alongside GPT 4.1 as the language model (LLM), which are both recommended for optimal performance.

• Latency should ideally remain under two seconds; users can monitor latency impacts based on prompt length and complexity.

Prompt Management and LLM Selection

• Retell offers minimal control over overarching system prompts but allows significant customization within user-defined prompts; keeping prompts under 3,000 tokens is advised.

• Exceeding 3,500 tokens incurs additional charges; strategies will be shared to manage token usage effectively as client demands grow.

Choosing the Right LLM

• While newer models like GPT 5 exist, GPT 4.1 remains highly effective due to its instruction-following capabilities and predictability.

• Default settings provide consistent latency without sacrificing speed during high traffic periods on OpenAI's servers.

Voice Configuration Options

• For voice settings, maintaining an LLM temperature at zero ensures structured output without excessive randomness in responses.

• Recommended voices include those from 11 Labs and Cartija due to their reliability; testing other voices is suggested but not prioritized initially.

This markdown file summarizes key points from the transcript regarding setting up an account and creating agents within the Retell platform. Each section highlights essential steps and considerations for users looking to optimize their experience.

Voice Settings and Functions Overview

Choosing the Right Voice Model

• The recommended voice model is 11 Labs Turbo V2 for English-only use cases; Turbo V 2.5 is available for multilingual needs, though its naming can be misleading.

• Voice speed adjustments are generally unnecessary unless specifically requested by clients; focus on voice temperature, which affects expressiveness and consistency.

Understanding Voice Temperature

• Higher voice temperatures lead to more variability in speech, making it sound "jumpy," while lower temperatures yield a monotone effect.

• A balance is typically found around 0.8 to 0.9, maintaining expression without excessive jumpiness.

Language and Transcriber Settings

• The default language setting for the transcriber is US English, with options for other accents (e.g., Indian, UK), but caution is advised when using multilingual settings due to potential errors.

• Multilingual support includes ten languages; however, sticking to a single language until proficiency is achieved is recommended.

Functionality Options in Call Management

• Default functions include ending calls and transferring them between agents fluent in different languages (e.g., English and Spanish).

• Appointment booking through Cal.com allows checking calendar availability directly within the system.

Advanced Features and Customization

• Sending SMS requires setup but can enhance communication capabilities during calls; extracting dynamic variables helps personalize interactions based on user input.

• Custom functions allow integration with external systems via API endpoints, providing flexibility beyond standard functionalities.

Knowledge Base Utilization

• Caution against using knowledge bases indiscriminately as they may introduce variability; structured information input into prompts yields better results than simple web scraping.

Speech Settings Optimization

• Responsiveness settings determine how quickly an agent reacts after a user finishes speaking; too quick responses can interrupt users' thoughts.

• Setting responsiveness at 0.98 or lower increases latency slightly but improves user experience by reducing interruptions during conversations.

This markdown file summarizes key insights from the transcript regarding voice settings, functionality options, knowledge base usage, and speech optimization techniques relevant to call management systems.

Understanding AI Interaction Settings

Background Noise Management

• The likelihood of background noise interrupting the agent's speech is a concern; starting settings are typically at 0.8 to balance responsiveness and clarity.

• In specific environments, like a doctor's office with children in the background, settings can be lowered to 0.6 to allow for user interruptions, though this may require longer user input.

User Experience Considerations

• Users may need to articulate full sentences to interrupt the AI, which could affect overall user experience; careful consideration is needed based on client expectations.

• Reminder messages are rarely necessary but can be set to trigger after one minute of silence if users become distracted.

Speech Normalization and Latency

• Speech normalization ensures that numbers and dates are presented clearly; enabling this feature may slightly increase latency but is generally advisable unless complaints arise.

• Pronunciation adjustments should only be made for unique names or terms relevant to the client’s business.

Transcription Settings Optimization

• Aggressive denoising options exist for transcription settings; however, if issues arise, reverting back to basic noise removal is recommended.

• Optimizing transcription for accuracy rather than speed helps capture complete user statements, particularly important for complex information like dates and numbers.

Vocabulary Specialization in Medical Use Cases

• Vocabulary specialization is crucial in medical contexts where certain terms may frequently be misinterpreted; adding hard-to-transcribe words can improve accuracy.

• This feature should only be used when there’s evidence of transcription errors related to specific terminology.

Call Settings Configuration

Voicemail Detection and Keypad Input

• Voicemail detection settings should remain active for outbound calls; agents can either hang up or leave a dynamic message based on detected voicemails.

• Keypad input detection is useful in phone tree scenarios but should remain off unless specifically required by the use case.

Call Duration Management

• Setting low thresholds for ending calls on silence (around one minute) prevents unnecessary costs from inactive calls.

• Maximum call duration should reflect realistic use cases—ideally double the expected maximum time—to prevent exploitation through prolonged calls.

Post Call Settings Utilization

• Post-call variables can enhance data collection by retrieving user information such as names during follow-up interactions.

• Providing multiple examples improves effectiveness when capturing optional fields like full names during post-call processes.

User Experience and Dynamic Variables in Voice Transcription

Importance of Accurate User Input

• The accuracy of transcriptions can be affected by the transcriber's ability to recognize names, even with the correct dialect settings. This highlights the need for careful consideration of user prompts.

• Asking users to spell their names instead of just stating them increases the likelihood of accurate transcription, emphasizing the importance of prompt design.

Model Selection for Accuracy

• Choosing the right model is crucial for handling dynamic variables; while GPT 4.1 mini suffices, opting for GPT 4.1 may enhance accuracy despite higher costs per call.

Security and Data Handling

• Retell ensures that sensitive information, such as personally identifiable information (PII), is not stored while still providing full transcripts and recordings.

• Users can customize what data should or shouldn't be stored, which is particularly important in enterprise-level applications.

Fallback Mechanisms

• In case a voice agent fails, selecting a fallback company like Cartija can ensure continuity in service without losing voice quality.

Dynamic Variables in Call Management

• Dynamic variables play a significant role during calls; they allow personalization based on user recognition and context.

• Implementing dynamic variables enables tailored greetings, enhancing user experience by addressing callers by name.

Webhook Settings and Integration

• Webhook settings are essential for directing end-of-call reports to appropriate destinations like GoHighLevel, ensuring efficient data management.

Building Effective Prompts

• When creating prompts, clarity in structure is vital; using markdown helps maintain organization and readability throughout the process.

Structure Components:

1. Role & Objective: Define the assistant's identity and primary goal clearly.

2. Personality & Context: Provide background about the client or user to tailor interactions effectively.

3. Instructions: Outline how tasks will be performed to guide agent behavior accurately.

Interaction Phases:

• Establishing stages within calls—like greetings and qualifying questions—ensures a smooth flow during interactions with users.

Example Interactions:

• Demonstrating expected dialogues helps agents understand how to maintain brand consistency while engaging with users effectively.

Consultation Process and Communication Techniques

Initial User Interaction

• The consultation process begins with a 30-minute session for qualified users, emphasizing a warm yet efficient communication style.

• Important context includes that the assistant is not a lawyer and cannot provide legal advice, with business hours set from 8:00 AM to 5:00 PM Eastern Time.

Dynamic Variables in Communication

• Introduction of dynamic variables such as current time formatted with double curly brackets, which helps personalize interactions by providing real-time context.

• Users' phone numbers can be dynamically included without asking directly, enhancing the agent's awareness during conversations.

Agent Awareness and User Familiarity

• The firm employs three lawyers specializing in different areas; their contact information is relevant when the user is known.

• A backend automation system allows for transferring users to specific lawyers based on prior interactions.

Effective Communication Strategies

• Instructions are organized into subsections using hashtags; effective communication tips have been refined over two years.

• Asking one question at a time improves clarity and ensures better understanding from both the user and the language model (LLM).

Enhancing Natural Conversations

• Emphasizing explicit instructions helps avoid confusion; natural filler words should be used sparingly to maintain conversation flow.

• Phone numbers should be read in natural groupings (3-3-4), while names must be spelled out clearly to ensure accurate understanding.

Spelling Out Information Clearly

• Dates should also be communicated clearly, spelling them out fully to avoid misinterpretation across languages.

• Attention must be paid to previously shared information by callers, ensuring that all relevant details are acknowledged during interactions.

How to Enhance Agent Versatility in Conversations

Importance of Avoiding Redundant Questions

• Acknowledge information provided by users to prevent asking the same questions multiple times, enhancing agent versatility and conversation fluidity.

Structuring Instructions for Agents

• Instructions encompass not only communication style but also available tools, starting with the "book appointment" tool, which should be clearly defined.

• Emphasize effective strategies for a smooth appointment booking process to minimize awkward exchanges regarding availability.

Stages of Conversation Flow

• Outline four key stages:

• Stage 1: Greeting

• Stage 2: Qualifying questions (proceed based on user qualification)

• Stage 3: Appointment booking

• Stage 4: Call closing.

Defining Qualifying Questions

• Consider where to specify qualifying questions; keeping stages simple is crucial as client interactions may introduce more complexities later.

• Create a separate section under instructions for qualifying questions, detailing what agents should ask and how to adapt based on user responses.

Key Qualifying Criteria

• Introduce a new subsection that outlines specific qualifying questions related to family law matters, emphasizing the need for adaptability during conversations.

• Highlight critical qualifying inquiries such as urgency of the situation and jurisdiction, which are essential for determining if legal assistance can be provided.

Qualification Criteria and Example Interactions

Overview of Qualification Criteria

• The location for qualification must be within New York State; failing any criteria results in disqualification from booking an appointment.

• If a caller does not meet the qualifications, the call may end or redirect to another law firm.

Example Interaction Setup

• The session begins with a qualified user interaction where an appointment is successfully booked.

• The digital assistant introduces itself as Grace, emphasizing transparency about being a voice agent to enhance user experience.

Importance of Transparency

• It's crucial for users to know they are interacting with an AI agent to avoid feelings of deception during the conversation.

• A standard greeting includes notifying that the call is recorded for quality purposes, which sets clear expectations.

Empathy in User Interactions

• When a user expresses distress (e.g., needing help with a divorce), the agent should respond empathetically, asking relevant questions about their situation.

• Caution is advised against relying solely on AI-generated prompts due to potential misunderstandings regarding platform capabilities.

Building Effective Prompts

• Manual prompt creation is recommended initially to develop skills and avoid confusion from complex AI-generated outputs.

• While brainstorming conversations can involve AI assistance, crafting prompts manually helps in understanding coherence and context better.

Coherence in Prompting

• Maintaining coherence throughout prompts is essential; if one question changes, it should be updated consistently across all related queries.

• An example question about urgency was deemed inappropriate for legal inquiries and removed to maintain relevance and coherence.

Dynamic Variables in Scheduling

• The conversation continues with qualifying details such as location and family involvement before scheduling consultations dynamically based on availability.

• Emphasis on dynamic variables allows flexibility in scheduling appointments rather than fixed dates, enhancing user experience.

Booking Appointments: A Step-by-Step Guide

Initial User Interaction

• The agent begins by asking the user, Sarah Johnson, to spell out her full name for confirmation. This step is crucial for ensuring accurate data entry.

• The agent confirms the activation of the booking appointment function and emphasizes the importance of explicitly showing how to use it.

Functionality and Availability

• The conversation includes a trigger for the booking function, with an emphasis on maintaining coherence in responses regarding availability.

• Dynamic variables are introduced to manage varying appointment times based on user input rather than hardcoding specific times.

Handling Edge Cases

• The agent prepares for potential issues by defining how to phrase availability when appointments cannot be booked successfully.

• An example interaction illustrates a successful booking process, highlighting the need for clear communication throughout.

Alternative Time Slots

• The discussion shifts to scenarios where users may not find initial time slots suitable. The agent must provide alternative options effectively.

• In this case, after confirming details about urgency and location, the agent presents additional time slots when initial suggestions do not work.

Testing and Adaptation

• After scheduling an appointment, testing methods are discussed. Two approaches include audio tests and chat tests for efficiency in identifying issues.

• Emphasis is placed on adapting interactions based on new information while ensuring that prompts remain coherent as they grow longer.

Streamlining Questions

• Strategies are shared for modifying questions within interactions to enhance clarity and relevance without losing essential context.

• By removing unnecessary questions from the flow, agents can streamline conversations while still gathering critical information from users.

Understanding Conversation Flow and Agent Versatility

Managing Conversation Prompts

• The conversation can be edited or resent, which replaces the previous messages and continues from the new point. This allows for a more streamlined interaction without asking repetitive questions.

• A preferred method involves using a "debug" button to regenerate multiple answers (up to 10). The goal is to ensure consistency across responses, aiming for at least nine identical answers out of ten.

Ensuring Consistency in Responses

• It's crucial for agents to maintain semantic consistency while respecting the flow of conversation. For example, variations in phrasing like "I'm sorry" versus "I'm really sorry" should not alter the intended meaning.

• When testing prompts, it's important to confirm that previously established facts (like the number of children in custody cases) are remembered correctly by the agent.

Appointment Booking Functionality

• The discussion transitions into setting up an appointment booking system, emphasizing the need for dynamic variables that reflect real-time availability before calls begin.

• Users must sign up for services like cow.com and connect their calendars during onboarding. This integration helps manage scheduling conflicts effectively.

Event Creation and Customization

• Creating events such as lawyer consultations requires personalization options. Users can add video conferencing links (e.g., Zoom), ensuring seamless client interactions.

• Availability settings allow users to define working hours and set minimum notice periods for appointments, enhancing scheduling efficiency.

Advanced Calendar Integration Features

• Users have various options regarding event frequency limits and buffer times between appointments, which help prevent scheduling overlaps.

• Dynamic variables can be utilized within calendar events to customize details based on user preferences or specific client needs.

This structured approach provides clarity on managing conversations with AI agents while also detailing how to effectively set up appointment systems tailored for legal consultations.

How to Effectively Use Cal.com for Scheduling

Integrating Event Types and API Keys

• The importance of avoiding unavailable booking times is highlighted, emphasizing Cal.com's versatility in integrating with various CRMs like HubSpot.

• To integrate an event type into the agent, users need the event type ID from the URL and an API key found in settings.

• Users are advised to create a master class API key that ideally never expires to prevent issues during scheduling.

Setting Up Time Zones and Descriptions

• When setting up appointments, it's crucial to specify the time zone accurately (e.g., American New York), especially when dealing with location-specific clients.

• A detailed function description is necessary; it should include how users will interact with the calendar system when booking appointments.

Formatting Dates and Parameters

• The time parameter must always be a full absolute date, even if users request relative dates. This ensures consistency in scheduling.

• All time entries should be set in the future, reinforcing clarity on time zones and required parameters for successful bookings.

User Interaction Flow

• Notes should summarize conversations briefly while dynamic variables like user email and names must be included for personalization.

• Emphasizes that while tools may seem intuitive, consistent usage requires skill developed over two years of iteration.

Booking Process Insights

• The process involves confirming user qualifications before proceeding with appointment bookings based on agreed times.

• Clear instructions on how to trigger functions effectively can enhance consistency in user interactions.

Dynamic Variables and Workflow Creation

• After confirming availability, ensure all bookings are made for future dates by checking against current availability dynamically.

• Introduction of n8n as a powerful automation platform for integrating systems efficiently; highlights its free trial offer for new users.

Onboarding and Cost Considerations

• Brief overview of n8n's onboarding process; emphasizes simplicity over cost-saving measures when building skills.

• Discusses pricing post-trial period ($25/month), suggesting that investing in effective tools is worthwhile compared to minor savings.

Understanding Availability Checks in Cal.com

Setting Up Availability Checks

• The process begins with checking availability in Cal.com before initiating a call, using dynamic variables to insert this information into a Retell agent.

• Key variables include the current time, check length, event type ID, time zone, and API key. It's important to avoid using dots in variable names to prevent errors.

• To reference the current time in N8N, use $now, which provides the date and time but may need adjustments for specific time zones like New York.

Configuring Time Zones and Dates

• Adjustments are made by adding hours or days to the current time; for instance, checking availability requires setting a minimum notice of two hours.

• Functions can be explored by typing . after a variable name; this reveals available functions that can be applied step-by-step.

Executing Steps and Validating Data

• After configuring the initial settings for availability checks (e.g., checking for seven days ahead), users must ensure all details like event type ID and API key are correctly filled out.

• Clicking "execute step" allows users to validate their inputs before proceeding further in the workflow.

Making HTTP Requests for Availability

• To retrieve actual availability from Cal.com, an HTTP request is initiated within N8N. This involves communicating with an external service via its API.

• Documentation is crucial; users should refer to docs.cal.com for guidance on how to interact with their API effectively.

Understanding Query Parameters and Headers

• Users need to identify necessary query parameters such as event type ID and start/end times when making requests. These parameters must be absolute dates.

• Authorization headers are also required during requests; ensuring these are set up correctly is essential for successful communication with the API.

API Key Management and Data Transformation

Secure API Key Handling

• The speaker discusses the importance of securely managing API keys, suggesting using header authorization to protect client privacy.

• Emphasizes the need to save the API key correctly without exposing it publicly, especially when working with clients.

Understanding API Responses

• The speaker explains how to interpret data returned from an API call, noting that it typically includes a list of available booking times.

• A sanity check is performed by comparing the returned data against actual calendar availability for accuracy.

Data Formatting for AI Processing

• Discusses transforming raw availability data into a more user-friendly format suitable for AI processing, highlighting the need for clarity in communication with AI systems.

• Introduces the concept of using LLM (Large Language Model) chains to convert technical data into natural language.

Prompt Creation and Variable Usage

• The process of creating prompts for LLM chains is outlined, emphasizing the use of variables to enhance flexibility and readability in outputs.

• Explains how JSON objects are utilized within prompts and introduces JSON.stringify as a method to make complex data more understandable.

Troubleshooting Common Errors

• Advises on handling errors during execution by focusing on red indicators in the interface, which signal issues needing resolution.

• Encourages self-troubleshooting rather than relying solely on AI assistance when encountering problems in model connections or configurations.

API Integration and Availability Management

Setting Sampling Temperature for API Responses

• The speaker discusses the importance of setting the sampling temperature to 0.0 for a specific client use case, emphasizing that randomness should be minimized.

• Different sampling temperatures are explained: 0.3 allows some randomness, while 0.7 introduces more variability; however, for this application, a raw output is preferred.

Creating an API Key from OpenAI

• Instructions are provided on how to create an API key in OpenAI by signing up and navigating to settings.

• After generating the key, it is pasted into the arena workflow credentials section to enable functionality.

Transforming Date Formats for Voice Agents

• The speaker explains how dates are transformed into a format that is easier for voice agents to communicate effectively.

• A dynamic variable setup is introduced to manage full availability over a week, with emphasis on formatting outputs as bullet points rather than paragraphs.

Configuring Soonest Availability Prompts

• The need for a separate prompt for soonest availability is highlighted; adjustments are made to ensure it captures only two upcoming times instead of all available slots.

• The speaker emphasizes the importance of offering varied time slots (morning and afternoon), ensuring that users receive practical options.

Troubleshooting JSON Data Issues

• An issue arises due to incorrect JSON notation; clarification on how data needs to be structured is provided.

• The solution involves using parentheses and selecting earlier notes correctly within the workflow setup.

Finalizing Appointment Outputs

• Adjustments are made in the code structure to ensure correct data retrieval from previous notes, focusing on appointment availability.

• Testing reveals issues with time zone discrepancies; solutions involve checking calendar settings and ensuring accurate representation of appointment times across different zones.

Event Type and Time Zone Adjustment

Changing Event Type and Time Zone

• The speaker discusses the need to change the event type from "Europe Lisbon" to "New York," emphasizing the importance of adjusting time zones for accurate scheduling.

• A mention of how time slots are affected by these changes, indicating that if it were Friday, there would be specific morning and afternoon times available.

Adapting Prompts for Use Cases

• The speaker reflects on adapting a previously successful prompt due to a shift in client location from Toronto to New York, highlighting the necessity of customization based on context.

• An issue arises with output errors linked to using an incorrect version (GPT 4.1 mini instead of 4.1), illustrating common challenges in tech adaptations.

Webhook Integration for Call Management

Setting Up Webhooks

• The discussion transitions into integrating webhooks for managing call availability before calls start, moving away from manual triggers.

• Emphasis is placed on using production URLs rather than test URLs when setting up webhooks, which is crucial for proper functionality.

Dynamic Variables and Response Handling

• The speaker explains the need for explicit paths in webhook configurations, particularly focusing on inbound call dynamic variables tailored for master class scenarios.

• It’s noted that most methods used will default to POST requests rather than GET requests unless specified otherwise.

Phone Number Setup and Inbound Call Agents

Purchasing Phone Numbers

• The process of purchasing phone numbers is outlined, including selecting area codes and payment requirements necessary beyond free trials.

Configuring Inbound Call Agents

• Details about assigning different agents for inbound/outbound calls are shared, along with options like SMS verification to prevent spam labeling.

Responding to Webhook Requests

Utilizing Webhook Responses

• The speaker highlights the importance of checking availability through inbound webhooks before answering calls, ensuring efficient communication flow.

JSON Structure Introduction

• An introduction to JSON format is provided as a method for crafting responses within webhook setups; it begins and ends with curly brackets containing key-value pairs formatted in quotation marks.

Understanding Inbound Call Handling

Setting Up the Inbound Call Response

• The process begins with an inbound call response, where a key is established to manage the call effectively.

• An override agent ID is specified to determine which agent will handle the call, allowing for targeted responses.

• Dynamic variables are introduced within an object structure, emphasizing the importance of matching variable names with those in prompts to avoid errors.

Configuring Availability Variables

• Two main dynamic variables are defined: "soonest availability" and "full availability," ensuring they align with previous data structures.

• The setup includes specific dates and times for availability, facilitating smooth communication between agents and users if initial options do not work.

Troubleshooting Call Issues

• After attempting to answer a call, troubleshooting steps involve checking for errors in execution logs and using debugging tools effectively.

• Attention is drawn to error messages (indicated by red highlights), guiding users on what needs fixing in their JSON code.

Validating JSON Structure

• Skills such as reading JSON and utilizing validators like JSON Lint are highlighted as essential for troubleshooting issues independently.

• A common mistake involves passing variable names instead of their values; understanding this distinction is crucial for successful calls.

Final Adjustments Before Making Calls

• Proper formatting of new lines in JSON requires using backslash N; this ensures that systems interpret data correctly without manual edits.

• The use of JSON.stringify helps format variables appropriately before making calls, preventing syntax-related issues from arising during execution.

Divorce Consultation Process

Initial Inquiry and Scheduling

• Alejo expresses the need for a divorce, indicating his current situation.

• The consultant confirms there are no children involved and schedules a 30-minute consultation for Alejo on Monday at 8:00 AM Eastern time.

• Alejo spells out his full name, "Aleho Kiduan," to ensure accurate documentation for the appointment.

Confirmation Details

• Alejo provides his email address as "aleho@amplifyvoice.ai" to receive confirmation of the appointment.

• There is a mention of an error in capturing the email, highlighting potential issues with automated systems in handling user information.

Understanding Inbound Agent Functionality

Features of the Inbound Agent

• The agent checks availability before answering calls and qualifies users based on their needs and location.

• It has awareness of its tools, including scheduling capabilities and qualifying questions relevant to family law matters.

Handling Non-Happy Paths

• Discussion about scenarios where users may not qualify for services due to being outside New York State or unrelated legal matters (e.g., car accidents).

• Observations on how some law firms might still conduct full intake even if consultations cannot be scheduled.

Iterating on User Experience

Importance of Manual Testing

• Emphasis on manual testing to ensure that essential functions like email collection work correctly during real interactions.

• Acknowledgment that the system failed to ask for email spelling, which could lead to errors in communication.

Information Gathering Strategies

• Recommendations for iterating through example conversations and refining booking tools based on user feedback.

• Suggestions include using default emails or phone numbers instead of requiring emails directly from users, enhancing overall user experience.

Next Steps in Client Management

Tech Stack and Telephony Integration

• Plans to discuss telephony integration further while ensuring leads are stored effectively within a CRM system as calls come in.

GoHighLevel Setup and Features Overview

Introduction to GoHighLevel

• The speaker introduces GoHighLevel, emphasizing its importance for remembering user interactions and setting up a path for incoming calls.

• A 14-day trial is mentioned, with an option to start at $97/month, transitioning to a $300 plan for unlimited sub-accounts.

Account Creation Process

• The process of creating an account includes setting a password and billing address while managing client numbers.

• The agency view is highlighted, focusing on the creation of sub-accounts and accessing shared snapshots.

Managing Contacts

• Key features such as calendars, contacts, opportunities, and automations are introduced; emphasis is placed on managing contacts effectively.

• The necessity of saving contact information (first name, email or phone number) is discussed to facilitate future interactions.

Automating Contact Management

• A manual addition of contacts is demonstrated; however, automation of this process is encouraged for efficiency.

• Each contact can have multiple fields added under general info like business name and address.

Calendar Integration

• Discussion on syncing appointments between GoHighLevel and Google Calendar; the need for two-way sync is emphasized.

• Steps to create team members in settings are outlined as part of calendar setup.

User Management in Calendars

• Instructions are provided on adding users directly from the admin account to manage calendars effectively.

• After creating a user, the ability to connect Google Calendar for appointment management is established.

Advanced Settings and Automation Triggers

• Advanced settings allow for triggers based on booked appointments; these automations enhance workflow efficiency within GoHighLevel.

Opportunities Management

• An overview of opportunities (pipelines), including stages like new lead and proposal sent, tailored specifically for law firms' needs.

Understanding the Law Firm Client Funnel

Overview of the Client Funnel Stages

• The client funnel for law firms is structured from unqualified leads to closed clients, tracking each stage meticulously.

• Initial stages include "New Lead Unqualified," followed by "Lead False Side but No Appointment," and then "Newly Qualified Appointment Booked."

• After an appointment, a consultation occurs with a lawyer, leading to potential closure of the client. If closed, payment reminders are automated for overdue invoices.

Managing Opportunities in the Pipeline

• Clients can access their lead status within the pipeline: New Lead Unqualified, No Appointment, Appointment Booked, Client Closed, and Payment Reminder.

• Demonstration of adding contacts to different stages in the pipeline shows practical application of managing opportunities effectively.

Automating Reminders and Workflows

Setting Up Automation Workflows

• GoHighLevel allows users to create intricate workflows starting from scratch; automation triggers can be set when an appointment is booked.

• An email reminder system can be established as part of these workflows to notify clients about upcoming appointments.

Customizing Email Notifications

• Users can customize email notifications using dynamic variables such as contact first name and appointment details (date/time). This personalization enhances communication effectiveness.

• Dynamic variables allow for flexibility in messaging; examples include inserting specific appointment times into reminders sent out before meetings occur.

Enhancing Communication Through Automation

Utilizing Dynamic Variables Effectively

• The use of dynamic variables simplifies communication by allowing personalized messages based on user data and scheduled events (e.g., appointment reminders).

• Options available through GoHighLevel enable sending SMS or emails easily while maintaining flexibility in how reminders are communicated to clients.

Importance of Workflow Flexibility

• Emphasis on adapting workflows according to user comfort levels and existing knowledge ensures that systems work efficiently for both service providers and clients.

• Saving actions within workflows guarantees that every time an appointment is booked, a reminder email will automatically be sent 24 hours prior to the event.

Finalizing Workflow Actions

Implementing Conditional Steps in Workflows

• Users have various options for setting conditions or waiting periods before actions are executed within workflows (e.g., sending reminders).

• Adding tags like "Appointment Booked" helps track client interactions more effectively throughout their journey in the funnel process. This tagging system aids in organizing client data systematically.

GoHighLevel Workflow Automation Overview

Moving Actions in GoHighLevel

• The speaker demonstrates how to move actions within the GoHighLevel platform, specifically placing an action immediately after an appointment is booked rather than waiting until 24 hours prior.

Integration with Google Calendar

• Appointments booked through Cal.com are automatically reflected in Google Calendar, showcasing seamless integration between platforms. The example given involves a lawyer consultation scheduled for 1 PM Lisbon time and 8 AM New York time.

Triggering Workflows

• Once an appointment is booked, a workflow triggers that adds a tag to the contact indicating "appointment booked." This highlights the flexibility of GoHighLevel's automation capabilities.

Pre-call and Post-call Processes

• The discussion shifts to integrating voice agents with GoHighLevel, focusing on pre-call processes (identifying users before connecting them to lawyers) and post-call processes (adding new users to the CRM).

Automation Paths in Workflows

• The speaker explains moving away from linear workflows by introducing conditional paths based on whether or not the user is known. This allows for tailored responses depending on user identification.

Integrating Webhooks for Contact Verification

Setting Up Inbound Webhooks

• A new workflow is created starting with an inbound webhook as a trigger. This setup will check if a contact exists in the CRM using their phone number.

Mapping References Requirement

• To successfully utilize webhooks, mapping references must be established. This includes sending specific information such as the contact's phone number during requests.

Checking Contact Existence

• The process involves checking if a contact exists based on their phone number. Two branches are created: one for when the contact does exist and another for when they do not.

Response Handling Based on User Status

• Regardless of whether the contact exists or not, both branches will respond similarly but will indicate true or false status regarding user recognition within GoHighLevel’s system.

This structured approach provides clarity on how GoHighLevel automates workflows and integrates various systems effectively while ensuring efficient communication management through webhooks and conditional logic.

Webhook Integration and Workflow Management in GoHighLevel

Setting Up the Webhook

• Mdn provides a resume URL for GoHighLevel, which will be used to communicate with the webhook after checking if a contact exists.

• The process involves passing not only the phone number but also a command to continue the workflow post-check. This is executed using native functions like "now" and "today."

Triggering and Cost Implications

• After checking if a contact exists, an inbound webhook is triggered, costing ￠1 per execution.

• For users on the $297/month plan, there’s an option for private integration that incurs additional costs but offers more control.

Handling Responses and Errors

• A POST request is set up to send data back to GoHighLevel; this includes whether or not a contact exists.

• If errors occur during execution (e.g., waiting for input), troubleshooting can be done by reviewing execution logs.

Adjusting Methods and Execution Flow

• An error was identified due to using GET instead of POST; changing this resolved the issue allowing workflows to proceed correctly.

• Once corrected, responses from GoHighLevel are received successfully, although further actions based on these responses need implementation.

Utilizing Contact Data

• The workflow now uses data such as contact ID and names to personalize interactions with users.

• An IF node is introduced to determine actions based on whether a contact exists or not, guiding subsequent steps in the workflow.

Creating New Contacts When Necessary

• If no existing contact is found, new contacts can be created using just their phone numbers through inbound hooks.

Call Handling and Appointment Booking Process

Overview of Call Handling

• The system is designed to check if a contact exists before proceeding with the call, allowing for flexibility in handling calls from unknown numbers.

• Upon executing the call, it waits briefly (around 4-5 seconds) before determining that the contact does not exist and proceeds to pick up the phone.

• The process involves minimal ringing (2-3 rings), which is deemed acceptable; however, adding more complexity could slow down response times.

Testing Call Scenarios

• The agent tests both scenarios: one where the user is unknown and another where they are known. This requires frequent updates to test numbers due to contact creation.

• During a live demo, an appointment is set up after confirming details about a divorce situation, showcasing how the system interacts with users.

Appointment Scheduling

• The agent gathers necessary information such as location and children involved in the case to facilitate scheduling a consultation.

• After confirming availability, the agent schedules an appointment for Wednesday at 9:00 AM Eastern time while ensuring all details are accurately recorded.

Data Management Post Call

• The need for capturing email addresses during calls was highlighted as an area for improvement in data collection processes.

• Acknowledgment that this demo agent still requires refinement; suggestions include implementing default emails or improving back-and-forth communication.

Integration with Calendar Systems

• After booking appointments, integration with Google Calendar allows users to see their scheduled meetings seamlessly.

• Users can view their appointments directly within their calendar applications, enhancing usability and accessibility of booked slots.

Future Enhancements and Considerations

• Discussion on making platforms like GoHighLevel more user-friendly by potentially integrating other tools like Chat Dash for better client interaction.

• Emphasis on understanding three phases of call management: pre-call (incoming call checks), actual call (interaction), and post-call (data analysis).

Integration of Systems with n8n

Setting Up Webhooks for Call Integration

• n8n is introduced as the integration platform to connect various systems before and after calls. A webhook is created for a POST request to handle end-of-call reports.

• During testing, three requests are generated upon ending a call: one for call start, one for call end (without analysis), and one containing the crucial call analysis data.

Handling Call Data

• The first webhook captures when a call starts, while the second captures when it ends but lacks important analysis data. The third webhook contains the necessary analysis.

• An "if" node or switch node is utilized to filter out unnecessary events, focusing only on the analyzed calls.

Updating CRM with Call Insights

• The goal is to send learned information from calls to GoHighLevel CRM, updating contact details such as names and emails that were previously missing.

• A post-call process involves setting up another inbound webhook to facilitate updates in GoHighLevel through HTTP requests.

Capturing Relevant User Information

• The system aims to capture essential user data like phone numbers and names during calls. This information will be used to update contacts effectively.

• It’s emphasized that full names should be parsed into first names for better handling within GoHighLevel.

Preparing for Post Call Analysis

• Before making test calls, it's crucial to clarify what user data will be captured—such as email formats and other relevant identifiers.

• Additional information collected includes appointment status (booked or not), which can help in understanding user interactions better.

Finalizing Data Capture Strategy

• Other key pieces of information include responses from users regarding their situations, which can inform lawyers about client needs during consultations.

How to Set Up a Lead Qualification System

Overview of the Process

• The discussion begins with outlining the answers needed for questions Q1, Q2, and Q3 regarding lead qualification and appointment booking.

• Emphasis is placed on using minimal tools to create an effective solution, recommending the GPT 4.1 model for data capture.

• A test call is initiated to demonstrate how the system works, highlighting the importance of capturing user responses accurately.

Call Simulation

• The digital assistant introduces itself and asks about the caller's situation regarding custody matters.

• The caller shares personal details about their custody issue, indicating a need for legal assistance in New York City.

• After discussing potential consultation times, the caller decides to call back later to schedule an appointment.

Data Handling Post-Call

• The focus shifts to analyzing post-call variables such as whether an appointment was booked or if the lead is qualified.

• Key data points are extracted from the webhook response including full name and qualification status of the user.

Updating Contact Information

• Steps are outlined for updating contact information based on call analysis results while maintaining consistent syntax throughout.

• It’s noted that this process can be replicated in different platforms like N8N or GoHighLevel.

Custom Fields Creation

• Discussion transitions into creating custom fields within a dashboard for better data management related to leads.

• Instructions are provided on adding new fields under settings in order to capture specific answers from callers effectively.

Creating Custom Fields and Automating Contact Management

Setting Up Custom Fields

• The discussion begins with the creation of custom fields for contact management, including full name, email, appointment booked (Boolean), and user qualified status.

• The "appointment booked" field is set as a single line input with a default value of false unless specified otherwise.

• After setting up custom fields, the next step involves creating a contact and ensuring that all relevant information updates in the CRM system (GoHighLevel).

Implementing Conditional Logic

• The speaker introduces conditional logic to manage different scenarios based on pipeline stages, starting with unqualified leads.

• If a lead is unqualified (user qualified = false), it triggers an update to the opportunity stage in the pipeline labeled as "unqualified."

• Similar conditions are established for other stages: new lead qualified but no appointment booked.

Managing Qualified Leads

• For leads that are qualified but have not booked an appointment, they will be updated to reflect their status in the pipeline.

• Execution logs confirm that opportunities are created or updated correctly based on user qualifications and appointment statuses.

Handling Appointments and Notifications

• A new branch is introduced for leads who are both qualified and have appointments booked; this updates their status accordingly.

• While some aspects can be controlled through automation, client interactions may still require manual adjustments by users within GoHighLevel.

Post Call Analysis Setup

• The importance of post-call analysis is emphasized for tracking follow-ups with clients who were qualified but did not book appointments.

• Internal notifications can be set up to alert users about issues during automation processes, ensuring effective monitoring of workflows.

• Advanced post-call analysis can include waiting periods before follow-ups via various communication channels like email or SMS.

Enhancing Client Interaction

• The final segment discusses improving client interaction by recognizing returning clients through personalized greetings based on their previous data.

How to Transfer Calls to Lawyers Efficiently

Setting Up Call Transfers

• The process begins with transferring calls to the appropriate lawyer, emphasizing the importance of adjusting settings when creating a new agent or duplicating an existing one.

• Key functions such as appointment booking and prompts are modified based on specific needs, ensuring that only necessary features remain active for efficient call handling.

• Two types of transfers are discussed: cold transfer (immediate handoff) and warm transfer (waiting for the recipient to answer), highlighting their operational differences.

• Custom messages are crafted for callers, personalizing interactions by addressing them by name and confirming their intent before transferring them to their lawyer.

• New custom fields are created in the system to store lawyers' names and phone numbers, facilitating accurate data management during call transfers.

Integration Considerations

• The integration of post-call analysis is suggested as a method for assigning lawyers based on caller needs, enhancing efficiency in client-lawyer matching.

• Flexibility in use cases is emphasized; while automation can be beneficial, it’s important not to delve too deeply into any single application prematurely.

Building Effective Prompts

• A step-by-step guide on building prompts is introduced, focusing on manual processes that enhance understanding and effectiveness in creating voice agents.

• The concept of a "meta prompt" is explained—this serves as a foundational template from which other prompts can be derived within cloud projects.

• Resources will be available in the Skool community for further learning about prompt creation techniques and best practices.

Optimizing AI Interactions

• Emphasis is placed on developing skills rather than merely adding tools; quality interactions require careful iteration and attention to detail in prompt crafting.

• Insights from OpenAI's research on GPT 4.1 are utilized to optimize prompts specifically designed for voice agents, ensuring they meet high standards of performance.

Structuring Knowledge Bases

• The structure of prompts mirrors that of effective voice agents; this includes personality traits and contextual information relevant to user interactions.

• A knowledge base document will be provided within the project framework, aiding users who wish to incorporate comprehensive information into their voice agent systems.

Creating Effective Voice Agents

Optimal Structure for Knowledge Base

• The speaker emphasizes the importance of a well-tested structure for knowledge bases, noting that it has proven effective without causing hallucinations in AI outputs.

• Best practices are encouraged to be modified according to individual use cases, allowing flexibility in adapting the system to specific needs.

Strategic Questioning in Prompt Creation

• When building prompts, the AI first asks strategic questions about objectives and client specifics before generating responses. This ensures tailored interactions based on user needs.

• Understanding the target audience is crucial; different demographics require varying tones—efficient for busy professionals versus empathetic for older users.

Stages of Prompt Development

• The process of creating new prompts involves four stages:

1. New prompt creation.

2. Auditing existing prompts.

3. Refinement through collaboration.

4. Testing key scenarios with transcripts as guides.

Context Gathering Before Drafting

• It’s essential to gather all relevant context before drafting a prompt to minimize revisions and ensure clarity in communication with users. Specific improvements should be targeted rather than general suggestions when upgrading prompts.

• Checking for conflicts or ambiguities within prompts is vital; the AI should challenge inconsistencies presented by users to enhance coherence and effectiveness in responses.

Crafting an Inbound Use Case Prompt

• A practical example is provided where an inbound agent is created for a law firm, focusing on transferring clients who already exist in their CRM directly to their lawyers while ensuring efficient communication during business hours.

• The speaker highlights that understanding your use case deeply influences how effectively you can design your agent, encouraging transparency with the AI about uncertainties regarding user needs and expectations.

Call Transfer Process for Amplified Family Law

Overview of Call Transfer Protocol

• The objective is to efficiently transfer clients to their assigned lawyers without wasting time, ensuring a streamlined process.

• Utilize existing client data, including lawyer names and contact information, to facilitate the transfer process effectively.

• Dynamic variables such as lawyer name and case type are essential for personalizing the call experience.

Execution of Call Transfers

• The digital receptionist must read instructions carefully to minimize errors; AI provides speed while human judgment ensures accuracy.

• Receptionist's role includes being professional, efficient, and courteous while keeping interactions brief and focused on completing transfers.

• The system retrieves client information before the call begins, allowing for immediate identification of the client's assigned lawyer.

Interaction Guidelines

• Follow a structured opening protocol: answer promptly with disclaimers and confirm the lawyer's identity before proceeding with the transfer.

• If a transfer fails or if the lawyer is unavailable, inform the client about business hours and ask if there’s anything urgent to note.

• Always redirect legal questions back to the lawyer instead of attempting to provide answers during calls.

Handling Various Scenarios

• Maintain clarity by asking one question at a time; keep responses light-hearted while focusing on connecting clients with their lawyers.

• After initiating a successful transfer, ensure that all necessary details are communicated clearly before ending the call.

Final Adjustments and Improvements

• Remove unnecessary reminders from prompts that may create bad habits; focus on coherent instructions that enhance overall performance.

Transfer Protocol and Dynamic Variables

Overview of Transfer Process

• The discussion begins with the introduction of a transfer number, emphasizing its importance in the protocol.

• A digital assistant introduces itself and informs the caller that the call is being recorded, offering to transfer them to their lawyer or inquire about other needs.

• The focus shifts to efficiency by removing filler words and unnecessary phrases to streamline communication.

Handling Transfers

• The conversation addresses ensuring correct terminology for "lawyer phone number" throughout the process.

• It is clarified that if a client declines a transfer, they will be informed that their lawyer will reach out during business hours.

Communication Standards

• Emphasis on showing rather than telling in initial greetings; examples are used to illustrate effective communication strategies.

• Suggestions are made for saving tokens by structuring conversations efficiently without repetitive greetings.

Dynamic Variable Management

Data Integration

• Discussion on integrating dynamic variables correctly into calls, including user and lawyer information.

• Importance of maintaining accurate notes during testing phases is highlighted, particularly regarding phone numbers used for testing.

Client Information Processing

• Steps are outlined for pulling necessary client data such as names and contact details before executing transfers.

• Post-call analysis involves categorizing lawyers based on specialties (e.g., divorce, child custody), which aids in directing clients appropriately.

Lawyer Selection Criteria

Specialization-Based Routing

• A selection mechanism is proposed where callers are routed based on their specific legal needs (divorce leads to Sarah Mitchell, etc.).

• The need for optional input fields is discussed, allowing flexibility when sufficient information isn't available.

This structured approach ensures clarity in understanding how dynamic variables and protocols enhance the efficiency of legal assistance through digital means.

Functionality Updates for LLM

Modifications to User Data Handling

• The LLM will output one of three options, ensuring all possibilities are covered. The function description is made optional.

• Emphasis on extracting the first name from a full name input; unnecessary text will be removed to streamline data handling.

• A similar process will be applied to retrieve the last name, ensuring consistency in user data extraction.

New Variables and Data Mapping

• Introduction of new variables: first name, last name, assigned lawyer, and assigned lawyer's phone number. These will be mapped correctly in the system.

• The need to refresh and update contact information in GoHighLevel (GCHL) with these new fields is highlighted.

Updating Contact Information

• Steps outlined for updating GCHL with first name and last name fields; users must ensure they make calls manually for accuracy.

• Creation of fields for assigned lawyer details is necessary; refreshing the interface allows visibility of new fields.

Call Reporting Enhancements

Importance of Call Details

• After qualifying a lead, it's crucial to send updated details including the assigned lawyer's information in post-call reports.

• Recording URLs are essential for tracking call history; they provide context during future interactions with clients.

Accessing Call Transcripts

• Instructions provided on locating recording URLs within JSON data received after calls. This includes navigating through transcripts effectively.

• Keeping track of both recording URLs and call summaries is emphasized as vital for client follow-ups.

Webhook Integration and Customization

Sending Additional Information

• New variables such as recording URL and call summary need to be integrated into GCHL updates via Webhook.

Custom AI Solutions

• Mentioned potential customization using AI agents for creating tailored call summaries, encouraging exploration beyond current topics.

Appointment Booking and Call Summary Integration

Creating Custom Fields for Call Management

• The process begins with booking an appointment, which includes a qualified recording URL and call summary. New custom fields will be created to store these variables.

• Two new custom fields are established: one for the recording URL and another for the most recent call summary. This setup allows retrieval of all notes related to a caller when they call back or when outbound calls are made.

• The current configuration is designed to remember only the most recent call. However, it is possible to retain multiple past calls by adding notes through an upgraded version using API instead of workflow triggers.

Implementing Custom Fields in Workflow

• Focus shifts to creating the two new custom fields within the appropriate folder, starting with a single line field for the recording URL.

• After saving changes, users can update contact information by scrolling down to find and utilize the newly created recording URL field during inbound webhook triggers.

Testing Contact Updates

• Upon triggering updates, it’s confirmed that the contact now has both a recording URL and a note summarizing the call details.

• Additional testing involves checking if contacts exist before adding further items like notes. A boolean variable is introduced to indicate whether a transfer was successful.

Dynamic Variables and Error Handling

Managing Agent IDs and Webhook Responses

• The agent ID is noted for routing scenarios effectively. The response format for webhooks needs adjustment as part of this management process.

• Dynamic variables are integrated into webhook responses, including known client agent IDs, ensuring easier maintenance of data flow.

Filling in Required Data Fields

• As dynamic variables are plugged into their respective fields, attention is drawn to ensuring consistency in naming conventions across all entries.

Final Testing and Identifying Errors

Executing Calls with Updated Information

• With everything published successfully, test calls are initiated to verify that all data points function correctly during interactions with clients.

Recognizing Issues in Data Handling

• An error is identified regarding inconsistent naming conventions within webhook configurations leading to incorrect data replacement in client records. This highlights the importance of maintaining accuracy throughout system setups.

Understanding Data Management and Troubleshooting in Call Systems

Importance of Careful Data Handling

• Emphasizes the need for caution when duplicating agents to avoid data loss, suggesting the use of checklists for verification.

• Discusses filling in missing data and highlights the importance of understanding the reasons behind data issues.

Executing Data Fixes

• Describes a process where previously inserted data can be retrieved to fill in gaps, noting that some fields may still require manual input.

• Explains how to save changes after updating an agent's end-of-call report, ensuring all necessary information is accessible.

Outbound Use Case: Managing Unpaid Invoices

• Introduces a scenario involving a client update on payment status, indicating follow-up actions for overdue invoices.

• Details the procedure for handling unsuccessful call transfers by sending emails to lawyers about clients needing callbacks.

Implementing Webhooks and Call Analysis

• Outlines setting up webhooks for post-call actions, including analyzing call outcomes based on transfer success or failure.

• Discusses using webhook responses to determine if further action is needed based on whether a lawyer has already engaged with a client.

Finalizing End-of-Call Reports

• Highlights the simplicity of creating email notifications within the system while ensuring proper formatting and activation settings are applied.

• Encourages flexibility in customizing end-of-call reports, allowing users to integrate various elements as needed.

How to Implement Call Summaries in Voice Agents

Setting Up Call Summary Variables

• The process begins with integrating a custom field for call summaries, allowing the system to retrieve and utilize this information during calls.

• When a call is received or made, the most recent call summary is passed along as a dynamic variable to enhance caller memory.

• A new dynamic variable named "most recent call summary" is created, ensuring that relevant data is accessible when calls come in.

• The setup involves organizing data retrieval from the weight node, which helps maintain clarity and efficiency in handling call summaries.

• It’s crucial to ensure that both known and unknown clients receive the correct call summary data through proper configuration.

Troubleshooting Dynamic Variables

• Issues may arise if incorrect notation (like JSON vs. full note names) is used; this can lead to undefined variables during testing.

• Errors should be viewed as opportunities for iteration; reviewing code and configurations can help identify mistakes without panic.

• Organizing agent IDs properly prevents future complications by making it clear where changes need to be made if issues occur.

Client Dashboard Integration

• The importance of an organized system becomes evident when managing disconnected calls; having access to previous summaries aids in continuity.

• Chat Dash is recommended for client-facing dashboards due to its user-friendly interface and effective presentation of information, despite being on the pricier side at $84/month after discounts.

Onboarding Clients with AI Services

• Clients can create their agency profiles within Chat Dash, customizing logos, domains, and themes while managing subaccounts effectively.

• Integrating external systems like Retell requires API keys for seamless operation of voice agents within the platform.

Final Steps in Agent Configuration

• After setting up an inbound agent for unknown clients, webhooks must be configured correctly for post-call reports to function smoothly.

• Once all settings are confirmed, users can log into their dashboard from a client perspective, providing insights into how services are delivered.

Dashboard Overview

Features of the Client Dashboard

• The dashboard provides a comprehensive view of all calls and analytics, allowing users to listen to call recordings and read transcripts.

• Users can access detailed analyses and summaries of conversations, presented in an easy-to-understand chart format.

Integrating Telephony Systems

Using Ring Central for Integration

• Ring Central is highlighted as a popular telephony provider among various options available for businesses.

• Integration involves forwarding calls from existing telephony systems to the voice agent, with options for setting business hours for call forwarding.

Setting Up Call Forwarding

• Users can configure incoming call rules to forward calls during specific hours or send them to voicemail if unanswered.

• Flexibility is emphasized; even if a different telephony platform is used, integration is possible through understanding the platform's capabilities.

Handling Unpaid Invoices

Workflow Triggering Calls

• A workflow checks daily for unpaid invoices over 30 days old, triggering follow-up calls based on this criterion.

• The process begins with a custom date reminder trigger that checks against the invoice date field.

Compliance Considerations

• Outbound calls are scheduled within compliance hours (Monday-Friday, 9 AM - 5 PM), adhering to legal time constraints based on time zones.

Executing Outbound Calls

Utilizing Webhooks for Communication

• The outbound call process will use webhooks to communicate between systems, specifically leveraging GoHighLevel as a timer mechanism.

• A new section will be created in the system setup focused on making outbound calls via HTTP requests.

Webhook Setup and API Integration

Configuring the Webhook

• The webhook is set up to trigger when an invoice is 30 days overdue, initiating a call at 9:00 AM.

• The process involves using n8n to create an HTTP request for the Retell AI API, specifically focusing on creating a phone call.

Authentication and API Key Management

• The authentication method required is Bearer token, which will be included in the header of the HTTP request.

• An API key must be generated from Retell's settings, ensuring it is named appropriately for easy identification later.

Required Parameters for API Call

• Essential parameters include 'from number', 'to number', and 'override agent ID'. These need to be formatted correctly in JSON.

• The body of the request will utilize JSON format instead of fields due to limitations encountered during setup.

Structuring Dynamic Variables

• Dynamic variables are created using expressions that allow for drag-and-drop functionality within n8n, requiring specific formatting.

• A nested object structure in JSON is necessary for sending multiple dynamic variables like user's first name and invoice date.

Finalizing the Call Setup

• Each dynamic variable must be separated by commas; however, the last entry should not have a trailing comma.

• If errors occur during testing (e.g., "no valid KYC"), steps for identity verification via QR code scanning with personal ID or passport are provided.

Getting Started with KYC and Outbound Prompts

Importance of Early Setup

• It is recommended to start the KYC (Know Your Customer) process as soon as you create your account to avoid future complications, as approval may not be immediate.

• Initiating the setup early allows for quicker approval, which is crucial for effective use of the system.

Testing Functionality

• A test call is performed to check for errors in the system, ensuring that everything functions correctly before proceeding.

• The focus is on confirming that data transmission works properly from the "go ahead" level.

Building Outbound Use Cases

• The discussion shifts towards creating an outbound use case specifically for payment reminders, emphasizing its effectiveness.

• Dynamic variables are introduced into the conversation design, which will enhance personalization and relevance during interactions.

Structuring Conversations

• Key dynamic variables include first names in greetings while other details like invoice dates are kept in context but not necessarily included in every interaction.

• The goal is to keep conversations straightforward, focusing on obtaining a clear answer regarding payment capabilities.

Call Management Strategies

• There’s a preference for triggering functions after calls rather than during them to reduce complexity and potential failures.

• Post-call actions allow more flexibility without latency concerns; however, confirming SMS receipt during calls can increase success rates.

Enhancing Payment Reminder Interactions

Dynamic Variables and User Engagement

• Incorporating dynamic variables such as payment amounts and outstanding invoices enhances user engagement by providing relevant information tailored to their situation.

Handling User Responses

• If users cannot pay immediately, offering follow-up options or payment plans can help maintain engagement and commitment.

Addressing Disputes Effectively

• In cases of disputes over payments, redirecting users to appropriate resources like lawyers ensures they receive proper assistance without overwhelming agents.

This structured approach provides clarity on key processes involved in setting up KYC procedures and developing effective outbound communication strategies.

Payment Reminder Call Process Overview

Initial Contact and Empathy

• The call begins with a payment reminder, emphasizing empathy while maintaining firmness regarding the outstanding balance.

• Acknowledgment of previous interactions is noted, highlighting the importance of continuity in conversations for effective communication.

Prompt Development and AI Utilization

• Discussion on modifying prompts to create concise and efficient communication strategies for client interactions.

• Emphasis on providing context to AI systems to enhance their understanding and response accuracy.

Structuring the Payment Reminder Script

• The script outlines the role of a payment reminder agent focused on collecting payments efficiently while confirming client readiness.

• Key attributes include being empathetic yet firm, ensuring brief and solution-oriented interactions.

Client Interaction Scenarios

• Various scenarios are discussed, including how to handle clients who can or cannot pay immediately.

• Custom messages are crafted for initial greetings, aiming for a friendly yet professional tone.

Follow-Up Procedures and Communication Standards

• Instructions are provided for follow-up actions if clients do not respond or confirm receipt of payment links.

• Guidelines emphasize clear communication standards when dealing with incorrect contacts or inquiries about invoices.

Client Payment Call Process

Initial Client Interaction

• The agent initiates a call regarding an outstanding invoice that is 30 days past due, emphasizing the need for prompt payment.

• The client expresses willingness to pay immediately if provided with a payment link, indicating readiness to resolve the issue.

Follow-Up Procedures

• It is suggested to standardize follow-up calls seven days after initial contact if clients indicate they will pay next week.

• Consistency in communication is highlighted as crucial for maintaining agent effectiveness and ensuring all necessary information is conveyed during calls.

Handling Client Queries

• If a client has already made a payment, agents should note this and maintain the introductory script for consistency.

• Agents are reminded to resend payment links if clients report not receiving them, ensuring effective communication throughout the process.

Functionality of Payment Link Sending

Tool Usage Instructions

• A new section on tool usage for sending payment links via SMS is introduced, detailing when to trigger this function based on client agreement.

SMS Options Overview

• Two options for sending SMS are discussed: using Retell's native function or utilizing n8n for custom functions. Each option has its pros and cons regarding reply capabilities and setup complexity.

Setup Considerations

• Setting up SMS through Toleo requires comprehensive information from clients and can take about a week; thus, early preparation is advised.

Creating Custom Functions

Function Triggering Details

• Emphasis on creating descriptive triggers for functions ensures better performance during execution. Clear descriptions help avoid unexpected issues during calls.

Webhook Configuration

• Every function begins with configuring a webhook; it’s essential to ensure successful message delivery by responding appropriately to webhooks.

Toleo vs. Retell Native Numbers

Cost and Setup Differences

• Toleo incurs monthly costs while offering more extensive setup requirements compared to Retell's native numbers.

Flexibility in Integration

• Toleo allows integration with various SMS services or CRM systems, providing flexibility in how messages are sent based on user needs.

Toyo vs. Tulio: Which Telephony Service to Choose?

Importance of Choosing the Right Telephony Service

• The decision to use Toyo involves identity verification and SMS registration, similar to Retell, but may lead to more complications.

• Building effective voice agents and prompts is essential; adding another service like Toyo can increase complexity but also offer integration possibilities.

• For beginners, Retell is recommended as it simplifies many processes by abstracting complexities.

Setting Up Telephony Functions

• Testing functions in a call setup is crucial; using test chat allows for monitoring function triggers effectively.

• Initial tests may result in errors due to incomplete setups, which are normal during the development phase.

Troubleshooting Call Functions

• If an error occurs during function execution, it can halt subsequent actions; reactivating functions may be necessary for continued operation.

• Successful calls require proper agent configuration and ensuring that all necessary data fields are filled correctly before making outbound calls.

Payment Link Integration

• During a call about outstanding invoices, sending a payment link is discussed as a way to facilitate immediate payments from clients.

• Two options for integrating payment links include using GoHighLevel with Stripe or creating charges through API keys for variable amounts owed by clients.

Error Handling and A2P Registration

• Errors related to unregistered numbers indicate the need for proper A2P (Application-to-Person) registration before sending SMS messages.

• Adding phone numbers into an approved A2P campaign allows successful message delivery, emphasizing the importance of compliance with messaging regulations.

Branding and Communication Strategies for Client Engagement

Importance of Proper Branding

• Emphasizes the necessity of having correct branding for agents to ensure they represent the brand effectively.

• Suggests adopting an empathetic approach by considering how clients perceive unexpected calls, particularly from lawyers regarding payments.

Enhancing Client Trust

• Discusses two options for verifying phone numbers to prevent calls from being marked as spam, which is crucial for outbound communication.

• Introduces a cost-effective solution called "number verifier" that charges $150 for cleaning 10 numbers, making it economical for businesses with multiple lines.

Workflow Implementation

• Outlines the process of triggering a call when an invoice is overdue, ensuring compliance with appropriate calling hours to avoid disturbing clients.

• Describes the technical setup involving webhooks and dynamic variables to facilitate sending payment links during calls.

Call Execution Example

• Provides a scripted example of an outbound call where the agent addresses an outstanding invoice and offers a payment link.

• Highlights effective communication strategies that make clients feel valued while addressing their debts without wasting time.

Post-call Analysis and Follow-up

• Mentions the need for an end-of-call report to analyze whether users agreed to pay during the call.

• Explains how to create this report using webhooks and emphasizes immediate responses based on user interactions during calls.

Custom Analysis Data Utilization

• Details how custom analysis data can be leveraged post-call to determine if users agreed to pay in full, enhancing follow-up strategies.

• Discusses ensuring data integrity by pulling information from reliable sources (webhooks), rather than potentially changing variables.

Voice AI System Development Insights

Workflow Design Considerations

• Emphasizes the importance of having a functional workflow that operates before, during, and after calls. Avoid over-engineering; focus on creating something effective.

• Discusses the process of sending emails based on user payment decisions, highlighting the need for clear communication with users regarding their payment status.

• Suggests implementing a waiting period (e.g., seven days) before following up with users who have not completed payments to avoid overwhelming them.

Client Collaboration and Flexibility

• Stresses the significance of client guidance in design decisions, advocating for visual tools that allow for iterative feedback and adjustments.

• Highlights the necessity of checking user payment status in CRM systems before making follow-up calls to ensure compliance and respect user preferences.

Compliance and Legal Considerations

• Underlines the legal requirement for obtaining user consent via checkboxes on forms or contracts to avoid illegal calls in regions like the US and Europe.

• Advises on time zone considerations when scheduling calls, ensuring compliance with local regulations about calling hours.

Spam Prevention Strategies

• Recommends using services like Number Verifier to clean contact lists and reduce spam risks associated with outbound calling campaigns.

• Provides guidelines on call limits per number (maximum 80 calls per day), emphasizing sustainable practices in outreach efforts.

Testing Framework for Voice Agents

• Introduces a structured testing framework involving happy paths (successful interactions) and unhappy paths (error scenarios), essential for refining voice agent performance.

• Encourages collaboration with team members (e.g., Claude) to develop specific test scripts that simulate real user interactions with voice agents.

This markdown file captures key insights from the transcript while providing timestamps for easy reference. Each bullet point is designed to be concise yet informative, facilitating efficient study and comprehension.

Voice Agent Development Insights

Understanding AI and Human Interaction

• The integration of human judgment with AI capabilities is crucial for effective voice agent performance, emphasizing the need for a balance between speed and decision-making.

• Payment requests should be handled separately to ensure clarity in communication; combining them can lead to confusion during interactions.

Iteration and Improvement Process

• Iterating on voice agent responses is essential, even if it seems tedious; this process significantly enhances the agent's effectiveness over time.

• Monitoring call history allows developers to identify issues in real-time, providing insights into function triggers and errors that may arise during calls.

Debugging and Error Management

• When encountering errors, it's important not to panic; focus on understanding the error messages rather than getting overwhelmed by technical jargon.

• Utilizing debugging tools can help assess how often an agent provides specific responses, allowing for adjustments based on performance metrics.

Maintaining Coherence in Prompts

• Keeping prompts concise is vital for coherence; lengthy prompts can complicate iterations and obscure specific behaviors of the agent.

• Modifying existing examples instead of adding new sections helps maintain clarity within prompts, preventing incoherences that could confuse the AI's learning process.

Effective Function Testing

• To confirm function reliability, rerunning conversations multiple times can validate whether functions trigger consistently as intended.

• Although testing may be time-consuming, thorough validation ensures that agents are equipped with necessary skills before deployment.

Selling Voice AI Systems

• Successfully selling voice AI systems hinges on clearly communicating their tangible benefits and ROI to potential clients.

• Building a network through personal connections (e.g., social media platforms like Instagram or LinkedIn) plays a significant role in establishing credibility when marketing these technologies.

Understanding the Importance of Your Network

The Role of a Warm Network

• A warm network consists of individuals who recognize you and your background, making outreach more effective.

• Many people hesitate to leverage their networks due to lack of experience or connections, but this mindset can hinder progress.

• Transitioning within familiar industries or roles is often easier than starting anew in an unfamiliar field.

Strategies for Client Acquisition

• Utilizing existing knowledge and connections can significantly reduce the time needed to secure your first client.

• Interest in a new industry is valid; however, leveraging contacts from your current network can lead to fruitful partnerships.

• If you have expertise in technology, consider collaborating with those who possess domain knowledge in other fields.

Choosing Your Niche

• It’s crucial to narrow down your target audience based on previous experiences while integrating new technological skills.

• Warm outreach is generally more effective than cold outreach, especially for those lacking extensive experience in sales techniques.

Identifying Reliable vs. Unreliable Niches

Characteristics of Reliable Niches

• Reliable niches typically feature high average order values and margins that allow businesses to invest comfortably in services.

• These niches are also characterized by ease of contact; many businesses may not yet be aware of emerging technologies like voice AI.

Understanding Unreliable Niches

• Unreliable niches often have low average order values (e.g., coffee shops), making it difficult for service providers to justify their fees.

• Low margins complicate business relationships as companies struggle financially, impacting their ability to hire external services effectively.

Decision-Making Challenges

• Larger companies present challenges due to hierarchical structures that make reaching decision-makers more complex.

• Smaller businesses may be easier to contact but might lack the financial resources necessary for investment.

Market Insights and Niche Selection

Understanding Market Dynamics

• The market is unpredictable, making it difficult for businesses to commit to monthly costs despite having high margins.

• Choosing a niche requires consideration of various factors, including personal connections and industry knowledge.

Complexity in Medical Practices

• Integrating medical practices is highly complex due to data privacy concerns and skepticism; prior experience in the industry can ease this process.

• Reliable niches identified include digital marketing agencies, law offices, home services (e.g., plumbing), and mortgage lending due to their urgency and profitability.

Unreliable Niches

• Food-related businesses, real estate (due to seasonality), and medical practices are considered unreliable niches because of integration challenges.

• If lacking experience or connections in these industries, it's advisable to avoid them initially as they may distract from achieving quick sales.

Focus on Quick Sales

• The primary goal should be securing the first sale quickly; practical experience is gained through action rather than theoretical learning.

• Best use cases involve clear ROI scenarios where leads express immediate interest, allowing for swift follow-up calls.

Importance of Value Over Trust

• Emphasizing value rather than trust is crucial; demonstrating how a service generates revenue is more effective than simply asking clients to trust you.

• Addressing existing business issues that are not currently being tackled can lead to successful sales opportunities.

Reactivating Old Leads

• Many businesses have old leads that remain uncontacted; reaching out can yield positive results as timing may now align with potential interest.

• A strategy could involve offering guarantees based on performance (e.g., earning a commission only if the client makes money), reducing perceived risk for the client.

Voice AI and Business Value

Focusing on Value Over Features

• Emphasizes the importance of concentrating on the value provided to clients rather than the features of voice AI technology. The primary concern for businesses is increasing revenue.

Speed to Lead

• Introduces "speed to lead," which refers to quickly contacting new leads, similar to leadless reactivation but focused on fresh prospects. This approach maximizes potential earnings by engaging warm leads promptly.

Importance of Timely Contact

• Highlights that immediate contact with leads significantly increases closing rates. A question posed during discussions is about the average time taken to reach out after a form submission, with proven statistics supporting quick follow-ups.

After-Hours Receptionist Benefits

• Discusses how an after-hours receptionist can ensure no calls are missed, enhancing client service without replacing existing staff. This guarantees capturing leads that could otherwise be lost due to missed calls.

Addressing Existing Issues

• Stresses that it's not about building trust but addressing unrecognized issues within businesses. Presenting solutions for problems they already face can effectively demonstrate ROI and ease understanding.

Research and Resources for Voice Agents

Exclusive Roadmap Access

• Offers access to exclusive research typically reserved for a community, providing a roadmap for building and selling voice agents aimed at guiding users through various industry opportunities.

Identifying Relevant Industries

• Encourages individuals to identify industries based on work experience, connections, and interests as part of decision-making in selecting niches or markets for their services.

Understanding Pain Points in Specific Niches

Niche Identification Process

• Outlines steps in identifying pain points before deciding on use cases. It emphasizes specificity in niches (e.g., types of real estate investors or dental practices), ensuring targeted solutions resonate with potential clients.

Examples of Specific Niches

• Provides examples such as medium-sized dental practices or reservation-only restaurants, illustrating the need for tailored approaches based on unique operational challenges faced by different sectors.

Common Challenges Across Niches

• Discusses common issues like lead generation versus lead qualification across various industries, highlighting that not all businesses struggle with generating leads; some may have sufficient demand but require better qualification processes.

Appointment Setting and Customer Support Challenges

Key Issues in Appointment Management

• Businesses face challenges in managing incoming calls, requiring effective appointment setting to ensure customer inquiries are addressed promptly.

• Onboarding processes can be cumbersome, especially in complex industries; a voice agent could streamline this process by guiding new employees through necessary documentation continuously.

Importance of Continuous Training

• Continuous training is essential for certain industries where employee knowledge must be regularly updated to maintain service quality.

Customer Support as a Critical Factor

• Effective customer support is crucial not just for client satisfaction but also for retention; 24/7 availability can significantly impact client decisions regarding staying with or leaving a business.

Identifying Target Personas and Pain Points

Defining Your Audience

• It's important to identify the specific persona that experiences the highest pain points your solutions can address.

Selecting Niche Focus Areas

• Choose one area of experience and two primary pain points to focus on when developing solutions tailored to your target audience.

Case Study: Day Spas

Common Challenges Faced by Day Spas

• Day spas often miss phone calls from potential clients after hours, leading to lost opportunities for bookings.

• Many spas struggle with maintaining a recurring client base, as clients may only book once without follow-up engagement strategies.

Solutions for Client Retention

• Current reliance on human receptionists is inadequate; they may not effectively reach out to all clients or manage appointments during busy times.

Proposed Use Cases and Solutions

Enhancing Business Value Through Technology

• The goal should be to ensure all calls are answered and appointments booked at any time, thereby increasing revenue potential for businesses.

Feedback Collection Strategies

• Implementing a voice agent that follows up with clients post-service can gather feedback and encourage rebooking through discounts, capitalizing on the client's immediate satisfaction.

Understanding Market Awareness and Education

Educating Clients on Value Proposition

• It’s vital not just to inform potential clients about features but also demonstrate the tangible value these technologies bring to their operations.

Final Steps: Clarity in Niche Selection

Packaging Your Solution Effectively

• After identifying your niche and use cases, focus on how you present your solution attractively to prospective clients.

Building Relationships

• Leverage existing networks when approaching initial clients; understanding who you know will help establish credibility in your offerings.

Understanding Client Communication and Objections

Importance of Clear Messaging

• Effective communication is crucial; it should convey specific benefits to prospects, such as who you help and the tangible ROI they can expect.

• Example provided: Helping single office family law practices convert leads after hours, preventing potential clients from going to competitors without needing extra staff.

Addressing Common Objections

• Anticipate objections related to hiring more staff or outsourcing; these are common concerns for potential clients.

• Engage with prospects about their experiences in hiring, emphasizing the challenges of finding suitable candidates in today's market.

Highlighting Unique Solutions

• Present your service as a distinct solution that alleviates pain points without the downsides of traditional methods like hiring or overseas outsourcing.

• Focus on results rather than features; for instance, emphasize how 24/7 phone answering leads to acquiring more clients at any time.

Tailoring Your Message

• Adapt language based on industry-specific pain points; specificity helps resonate with the right audience.

• The goal is not mass appeal but connecting with individuals who recognize your service as a solution to their needs.

Leveraging Your Network

• Emphasize the importance of tapping into your warm network for initial sales opportunities rather than seeking numerous leads.

• A simple outreach message can be effective: ask acquaintances if they know business owners who could benefit from your services without pitching directly.

Building Connections Through Introductions

• Request introductions through mutual contacts by framing it as gathering information rather than making a pitch.

• Suggest using Zoom or in-person meetings for discovery calls to foster trust and encourage open sharing of insights.

Expanding Your Network Strategically

• If lacking connections, consider attending conferences or business groups relevant to your niche (e.g., BNI, chamber of commerce).

• While this approach requires investment in time and money, it can effectively build networks and generate client opportunities.

Discovery Calls and Networking Strategies

Building Your Agency Network

• Emphasizes the importance of leveraging your network after initial agency setup, suggesting calls with contacts to explore opportunities.

• Highlights the goal of discovery meetings: understanding client situations, qualifying leads, and generating excitement without overpromising.

• Advises focusing on the client's needs during discussions; redirect inquiries about your services back to their experiences and challenges.

Key Questions for Discovery Meetings

• Suggests asking about current call handling processes, missed calls, and intake procedures to gather valuable insights into client operations.

• Encourages exploration of clients' marketing and sales processes to identify how they attract customers and enhance client lifetime value.

Listening for Opportunities

• Stresses the importance of identifying pain points and buying signals during conversations; avoid pitching too early in the process.

• Recommends painting a scenario for clients about life without their current issues, fostering a human connection rather than a transactional one.

Transitioning to Demos

• Discusses the potential transition to demos in later meetings but cautions against doing so prematurely if you lack sufficient information about the client.

• Describes demos as an opportunity for shared experience rather than a hard sell; focus on creating memorable moments even if technical aspects fail.

Defining Your Niche

• Encourages setting up follow-up meetings based on initial discussions while still exploring various niches within your network.

• Advises finding what excites you personally about potential niches since emotional engagement is crucial for selling new technologies effectively.

Pricing Strategies Overview

• Introduces different pricing models including one-time setup fees or partial payments upfront as part of establishing service agreements.

Pricing Models for Voice Agent Integration

Initial Setup and Recurring Fees

• The pricing model should include a setup fee, which covers building, integrating, launching, and post-launch improvements based on real-world feedback.

• A monthly improvement fee can be introduced where the service provider actively listens to calls and enhances the voice agent's performance.

• Businesses may prefer scrapping the setup fee due to upfront investment concerns; offering a no-commitment monthly plan could alleviate this hesitation.

Usage-Based Pricing

• Consider implementing usage-based pricing per call or minute while ensuring healthy profit margins; avoid competing with low-cost services that offer poor quality.

• Competing against cheap services is not advisable as it undermines the value of high-quality systems that are reliable and seamlessly integrated.

Success-Based Pricing Challenges

• Initially avoid success-based pricing models as they can lead to financial losses if not executed properly; these models depend heavily on client sales processes.

• Performance-based models mean "if you make money, we make money," but early-stage performance may not yield significant results.

Suggested Pricing Strategies

• Two suggested options:

• Charge a one-time fee of $1,000-$2,000 including three months of support post-launch.

• Alternatively, implement a no setup fee with a charge of $0.50 per minute.

Long-Term Profitability Models

• As more clients are acquired, the per-minute margin can become profitable over time despite initial low earnings from first clients.

• After gaining experience with initial clients, consider long-term pricing strategies such as $497/month plus $0.40/minute or charging extra for CRM integrations.

Flexibility in Pricing Structures

• Offer discounts for upfront payments (e.g., 33% off for annual payment), while maintaining consistent per-minute rates to attract price-sensitive niches.

• Emphasize aligning incentives between your business and clients through flexible pricing structures that ensure profitability while being attractive to customers.

Pricing Strategies for Voice AI Solutions

Understanding Commitment and Pricing Models

• The speaker discusses a yearly commitment model that can save clients $2,000 annually, emphasizing the importance of upfront payment to facilitate discussions about pricing.

• Flexibility in pricing is crucial; instead of overwhelming potential clients with multiple models, focus on one core offer while exploring their needs during discovery calls.

Success-Based Pricing Challenges

• A success-based pricing model at $10 per appointment requires a fully integrated system to track client interactions effectively; otherwise, it may not be scalable.

• The effectiveness of cold calling with AI is questioned; while some cases succeed (e.g., offering free samples), general B2B cold calling is discouraged due to high failure rates.

Hybrid Pricing Models

• Stable pricing models are often hybrid, combining monthly fees with additional charges based on usage. This approach allows for increased revenue while being perceived as fair by clients.

• To determine appropriate pricing, businesses should analyze costs and industry benchmarks. Researching competitors' prices for similar services can provide valuable insights.

Leveraging AI for Market Research

• The speaker suggests using AI tools like Claude to research what law firms pay for call intake and payment reminder services to better position their voice AI solution.

• While AI can conduct deep research quickly, human judgment remains essential in understanding nuanced situations that technology may overlook.

Market Positioning and Competitive Landscape

• The market for voice AI solutions is becoming competitive; aiming for premium service tiers rather than competing solely on price is recommended.

• Services priced between $200-$600/month typically include comprehensive features such as intake qualification and appointment scheduling, which align with the offerings being developed.

Cost Comparisons in Service Offerings

• Premium services costing $1,000-$2,000/month provide dedicated teams familiar with specific client needs and protocols across various legal fields.

• Analyzing incumbent human answering services reveals average costs around $3.50 per minute; the goal is to offer more efficient solutions at lower rates.

Pricing Strategies and Value Alignment in Business

Importance of Positioning and Framing

• Effective positioning and framing are crucial for success; competition should focus on value rather than price.

• Flexibility in pricing models is important, but it should not lead to indecision; come prepared with a clear offer.

• Aligning incentives through negotiation can help close deals by adapting to the prospect's existing service preferences.

Focus on Benefits Over Features

• Emphasizing benefits rather than features is essential; competing on features leads to neutrality while clients may feel uncertain about AI solutions.

• Understanding the client's current situation, such as call rates and missed calls, helps communicate relevant benefits effectively.

Strategic Pricing Models

• ROI-focused messaging highlights financial losses from missed calls instead of just promoting services.

• Competing on price with new technologies often results in unfavorable outcomes; consider hybrid pricing models that combine monthly fees with per-call or per-minute charges.

Incentive Alignment Challenges

• Including a set number of calls or minutes can misalign client incentives, leading to complications when clients want to stop services after reaching limits.

• Offering one-time bonuses (e.g., 1,000 free calls upon sign-up) can be attractive without creating ongoing incentive issues.

Performance-Based Pricing Insights

• Value-based pricing linked to qualified leads delivered aligns payment with outcomes, which is appealing for many businesses.

• For collection services, charging a percentage of payments collected ties costs directly to successful outcomes.

Leveraging Technology for Presentations

• Positioning payment reminder services as valuable add-ons can attract law firms that currently handle these tasks internally.

• Utilizing tools like Gamma for presentations allows effective communication of offers and research without extensive time investment.

Creating Presentations with Gamma

Overview of Gamma's Features

• The speaker introduces Gamma, highlighting its ease of use by allowing users to log in with a Google account and create presentations or web pages.

• Emphasis on minimal text in presentations; the speaker prefers concise content and selects a memorable theme for better engagement.

• A comparison is made between creating a presentation and a website, noting that each format conveys different energies and pacing.

Presentation Customization

• The speaker discusses generating simple presentations with a maximum of six slides, maintaining consistency in themes and settings across formats.

• Features like animations are highlighted as enhancing the presentation experience, making it more dynamic and engaging.

Importance of Use Case Validation

• The speaker stresses the significance of validating use cases during initial discovery calls to ensure realistic expectations from potential clients.

• Mentioned is the challenge of presenting inflated statistics that may not resonate with smaller firms, emphasizing the need for industry knowledge.

Sales Strategy Insights

• The discussion shifts to focusing on specific use cases rather than broad industries; an AI sales intake agent example illustrates this approach effectively.

• The speaker shares insights about their own product offerings, including features like call transfers and CRM integration tailored for niche markets.

Pricing Structure Breakdown

• An overview of pricing is provided: no setup fee at $400/month plus additional charges per minute for calls.

• Details on various plans are shared, including discounts for longer commitments and added features such as dedicated support channels.

Conclusion on Pricing Models

• The importance of understanding diverse pricing models is emphasized; effective communication about what's included can enhance client relationships.

Understanding Pricing Models and Resources

Commitment Levels and Resource Sharing

• The speaker emphasizes the importance of different commitment levels (e.g., six-month, twelve-month) in pricing models, encouraging listeners to create their own versions based on a provided template.

• A resource is introduced that allows users to generate artifacts or links for prospects using Claude, which can be transformed into websites or presentations.

Discovery Call Preparation

• The speaker discusses the significance of discovery calls within the sales process, highlighting resources available for effective preparation.

• Access to a GPT tool is promised, aimed at helping users prepare questions related to pricing and client interactions without gatekeeping information.

Utilizing AI Tools for Client Engagement

• The GPT tool can assist in formulating relevant questions during discovery calls, enhancing understanding of potential clients' needs.

• Emphasizes the concept that "preparation meets opportunity," suggesting that being well-prepared increases chances of success in client engagements.

Financial Metrics and Cost Analysis

• Key metrics are discussed such as client value, missed calls, and conversion rates; these help quantify potential revenue from new clients.

• The speaker illustrates how capturing additional clients correlates with financial gains, making a case for investing in services that yield high returns.

Operational Costs Breakdown

• Detailed breakdown of operational costs associated with running a voice AI agency is provided, including specific service costs like Retell and GHL subscriptions.

• Suggestion to allocate around $1,000 monthly for initial business setup costs over three months is made to ensure adequate funding before generating profit.

How to Effectively Close Deals and Deliver Value

Setting Up Demos and Presenting Pricing

• Discusses the importance of setting up a demo meeting after initial contact, emphasizing that it's not just about the demo itself but also how to present pricing and handle objections effectively.

Closing Clients and Delivery Checklists

• Highlights the significance of having a checklist for delivering services once a client is closed, noting that while it may vary by client, there are standard elements that should be included.

Leveraging Your Network for Support

• Suggests reaching out to your warm network for support, particularly looking for individuals with domain expertise who can complement your technology skills.

The Importance of Persistence in Early Stages

• Emphasizes the critical nature of the first three months in building a business, where persistence is key to avoid burning contacts or giving up prematurely.

Framework of Support for Business Acceleration

• Introduces a three-level framework designed to accelerate business growth through structured support via Alex Hormozi's Platform Skool. This includes resources like YouTube videos and templates aimed at helping users build their offerings effectively.

Accessing Build Hours and Community Interaction

• Describes "build hours" as an essential part of support where clients receive hands-on help from coaches in group settings. This interaction fosters community learning and development.

Affordable Support Options

• Mentions that standard support is available at $37 per month, providing access to coaching and community interactions without gatekeeping knowledge.

Premium Support Features

• Outlines premium features including sales hours focused on selling strategies, case studies from past clients, and exclusive workshops aimed at refining sales techniques.

Learning Through Community Engagement

• Explains how community engagement allows members to learn collectively; when one person solves an issue, everyone benefits from shared knowledge through recorded sessions.

VIP Tier Benefits and Sales Process Insights

Overview of VIP Tier Offerings

• The VIP tier includes a $400 hourly consultation, a 30-minute coaching call with co-founders, and access to various sales and development coaching calls each month.

• Members will have premium access to classroom modules, recordings, and future calendar events. An exclusive mastermind group with private chat for coaches and fellow members is also included.

• Free access to in-person events is offered, including past events in New York and upcoming ones planned for Portugal in 2026.

Key Strategies for Closing Sales

• The speaker emphasizes the importance of providing an experience during sales meetings rather than just pitching prices. This involves three key strategies: giving an experience, making it visual, and asking questions.

• Preparation is crucial; tailored slides or websites should be created based on prior conversations with prospects to address their specific needs.

Structuring the Sales Presentation

• Presentations should start by addressing pain points directly followed by solutions. Simplified explanations are essential since clients may not understand technical jargon like "voice AI."

• Include expected results from the solution using case studies while clarifying that these are potential outcomes rather than personal achievements if no previous results exist.

Conducting Effective Live Demos

• Begin live demos with a clear statement of how the agent helps specific client types achieve desired outcomes without major pain points.

• Use realistic scenarios that prospects can relate to during demonstrations. Focus on two or three key features that provide direct value instead of overwhelming them with all functionalities.

Handling Challenges During Demos

• If issues arise during live demos (which is common), educate clients about potential delivery timelines due to complexities involved in ensuring consistent performance.

• Demonstrate both easy and challenging responses during the demo to showcase the agent's capabilities effectively. Prepare for interruptions or edge cases as part of the presentation strategy.

Live Demos and Sales Strategies

Importance of Live Demos

• The Build Stage emphasizes the significance of scheduling build hour demos to prepare effectively, run the demo, and receive feedback on enhancing sales effectiveness.

• Recording a demo can also be beneficial; however, live demos provide a unique experience that is crucial for engaging prospects.

Preparing for Demos

• If connectivity issues are anticipated during in-person meetings or conferences, pre-recording a live demo is advisable. This allows prospects to view the interaction while still experiencing the agent's capabilities.

• It's essential to guide prospects through the standard process by having them call an agent after your initial call, ensuring they understand how interactions flow.

Utilizing Call Reports and Dashboards

• After calls, Retell sends an end-of-call report. Using tools like ChatDash can visually present call outcomes, including recordings and transcripts, making abstract experiences tangible.

• Clients benefit from seeing analytics and conversation summaries through dashboards that enhance transparency regarding lead interactions.

Engaging Prospects Post-Demo

• Following a demo, it's important to discuss next steps with clients based on their needs. Asking questions keeps conversations focused on their experiences rather than unrelated concerns.

• Addressing potential objections about AI interactions involves framing benefits clearly—emphasizing that leads calling need immediate assistance which AI can reliably provide.

Handling Objections Effectively

• When faced with objections about AI usage in customer service, highlight studies showing that people prioritize problem-solving over human interaction preferences.

• Control the conversation by asking how ready they are to move forward on a scale from 1 to 10. This approach helps uncover hidden objections without guessing.

Closing Techniques

• If prospects express hesitance (e.g., rating below 10), continue probing for additional concerns while steering discussions back towards tangible benefits.

• Regardless of readiness to close immediately, always schedule follow-up meetings during current discussions to maintain momentum in negotiations.

Client Engagement and Project Management

Availability and Meeting Structure

• The speaker emphasizes their availability on Tuesdays or Thursdays for meetings, suggesting that scheduling should align with this.

• It's common for clients to require three to four meetings to close a deal; however, more niche offerings may shorten this timeline.

Contract Essentials

• A scope of work document is crucial; it outlines what the project includes and excludes, differing from the formal contract.

• Contracts should be straightforward—avoid lengthy documents. Key elements include pricing, payment terms, support details, and termination clauses.

• Payment collection methods (e.g., Stripe links or bank transfers) should be clearly defined in the agreement.

Kickoff Call Importance

• After signing the contract, a kickoff call is essential to clarify workflows and gather necessary access information.

• Clients should walk through their processes explicitly during the call to ensure clarity on how they operate.

Setting Expectations

• Establishing timelines is critical; unexpected issues (like CRM access delays) can affect project schedules.

• Communicate any potential delays early and maintain flexibility while being transparent about progress.

Defining Success Metrics

• Clear success metrics must be established so both parties know when a project is complete. This prevents ongoing changes from extending timelines unnecessarily.

• Educating clients on testing importance helps manage expectations regarding project duration and outcomes.

Training Requirements

• Understanding team roles (e.g., receptionists' involvement in new processes) is vital for effective training post-deployment.

• Anticipate potential employee resistance to AI implementations; clear communication about changes can mitigate frustration.

Pre-Building Steps

• Before sending an agent for client feedback, five key tasks must be completed: creating prompts, integrating with CRM systems, setting up post-call automations, ensuring telephony integration, and systematic testing across various scenarios.

Testing and Launching AI Agents

Importance of Testing AI Agents

• Emphasizes the necessity of testing all possible paths for AI agents, especially after making changes to prompts. This ensures comprehensive coverage in testing.

• Suggests batching changes before retesting all paths, which streamlines the process and maintains organization.

Client Interaction During Development

• Recommends setting up a dashboard for clients to review progress at the halfway point of a project timeline, enhancing transparency and collaboration.

• Highlights the importance of managing client expectations during a pre-launch meeting, preparing them for potential issues post-launch.

Post-Launch Refinements

• Discusses launching an agent for a limited time to gather real-world data, which is crucial for refining performance based on actual usage.

• Advises allowing three months for post-launch refinements, particularly with initial clients, to ensure optimal functionality and satisfaction.

Transitioning Roles: Builder to Consultant

• Stresses the need to shift from a builder mindset to that of a consultant as real-world data comes in; this adaptability is key in maintaining client relationships.

• Mentions that keeping clients happy is paramount and suggests using checklists throughout the process to ensure thoroughness.

Comprehensive Resources Available

• Introduces a detailed checklist covering everything from kickoff calls to post-launch refinements, designed as a master tool for success.

• Notes that not all checklist items will be relevant for every client (e.g., outbound agents), allowing flexibility based on specific needs.

Accessing Resources and Automation Tools

• Explains how resources can be exported from Retell accounts and imported into new setups easily, facilitating smoother transitions between projects.

• Provides guidance on accessing GoHighLevel automations through shared snapshots, emphasizing ease of use within sub-account structures.

How to Set Up Your Voice AI Agency

Importing and Setting Up Snapshots

• To sign up for a 14-day free trial, follow the provided link. After signing up, you can import a Snapchat into your account by selecting "yes" when prompted.

• When creating a new sub-account with the imported snapshot, ensure to update the webhook settings as they will differ from the original setup.

Configuring Automations

• Download workflows from n8n and import them into your system. It's crucial to change webhooks specific to your setup during this process.

• Create custom fields in GoHighLevel for each automation. This flexibility is essential for ensuring that all elements function correctly within your agency's framework.

Understanding Webhook Settings

• Update API credentials and HTTP calls for services like Twilio. Each webhook URL must be tailored to fit your domain and service requirements.

• Ensure that every agent imported has its webhook settings updated accordingly, particularly for end-of-call reports and appointment bookings.

Master Prompt Access

• You will receive access to a master prompt document that aids in creating prompts for various projects. Utilize these resources effectively to launch your agency.

Key Concepts of Building a Voice AI Agency

Market Timing and Use Cases

• The current market presents an excellent opportunity for establishing a voice AI agency, especially targeting early majority clients.

• Discussed the anatomy of voice AI agents, focusing on inbound versus outbound use cases inspired by real client scenarios.

Tech Stack Considerations

• Identified suitable platforms (Retool, n8n, GoHighLevel, Cal.com) necessary for building effective solutions tailored to client needs.

Sales Strategy Insights

• Emphasized leveraging personal networks as the quickest route to acquiring initial clients; provided questions to refine niche targeting.

Niche Selection and Pricing Strategies

Reliable vs Unreliable Niches

• Offered guidelines on choosing reliable niches while considering potential risks associated with unreliable ones.

Best Use Cases

• Recommended starting points such as lead reactivation strategies and after-hours receptionist services due to their clear value propositions.

Pricing Models

• Discussed various pricing models available for agencies; highlighted how justifying prices through estimated client savings can enhance sales pitches.

Operational Guidelines for New Agencies

Client Acquisition Goals

• Suggested aiming for first client acquisition within three months with an investment strategy focused on achieving profitability quickly.

Community Engagement

• Encouraged participation in community forums like Skool during free trials; offered onboarding support through coaching sessions.

How to Effectively Launch and Improve Your Client Projects

Steps for Successful Project Launch

• Emphasizes the importance of making the project visual and engaging, which increases the likelihood of client buy-in.

• Discusses the process of delivering signed contracts, conducting kickoff calls, building projects, and launching them effectively.

• Highlights the significance of post-launch reflection to gather insights that can enhance future projects.

Community Engagement and Acknowledgments

• Encourages connecting on Skool and LinkedIn for further engagement, noting a preference for activity on Skool over other social platforms.

• Expresses gratitude towards co-founders Paige and Kevin, acknowledging their contributions to the success achieved.