How to Build a 24/7 Legal Answering Service Using AI

If you run a law firm, missing after-hours calls could be costing you hundreds of thousands in lost revenue. Imagine if you could delegate that critical first contact to a custom-trained AI that provides instant, professional responses 24/7, captures qualified leads, and books appointments while you sleep.

In this comprehensive guide, I'll show you exactly how to build a powerful multi-agent AI receptionist that connects to your business systems, understands complex legal inquiries, and handles everything from intake to appointment scheduling. Using our "Goal Line Legal" sports law firm as a demo, this AI assistant will transform how you capture and convert prospects.

AI Legal Receptionist Key Features

We are creating a sophisticated AI system that goes beyond simple chatbots. This assistant will:

Understand Natural Conversation: Callers can speak naturally about their legal needs without rigid menu systems or keywords.

Intelligent Routing: A "Referee" agent analyzes caller intent and routes them to specialized AI assistants.

Professional Knowledge Base: Answers common legal questions using your firm's actual information and policies.

Lead Capture & Qualification: Systematically collects contact information and case details for new prospects.

Appointment Booking: Handles consultation scheduling with intelligent date/time processing.

Seamless Human Handoffs: Escalates complex cases to human attorneys with complete context.

24/7 Availability: Never miss a potential client, regardless of time or day.

Business Intelligence: All interactions automatically flow into your CRM and email systems.

The Technology Stack

This powerful solution combines three key platforms, each serving a specific role:

Voiceflow: The AI Brain. We use Voiceflow's multi-agent system to design sophisticated conversation flows and manage the firm's Knowledge Base.

Make.com: The Business Logic. Make.com handles all the backend integrations, connecting our AI to Google Sheets, email systems, and other business tools.

Google Workspace: The Data Layer. Google Sheets serves as our CRM database, while Gmail handles professional email notifications.

How to Build Your AI Law Firm Receptionist

Prerequisites & Setup

Before we begin, ensure you have accounts for:

• A Voiceflow Account (Pro plan recommended)
• A Make.com Account
• A Google Workspace Account
• Access to your law firm's knowledge base materials

Want a head start? All the code, templates, and Make.com scenarios for this entire tutorial are available at the end of this post.

Phase 1: Building Your Knowledge Base Foundation

Step 1: Preparing Your Legal Knowledge Base

The foundation of any intelligent legal assistant is accurate, comprehensive information. In Voiceflow, navigate to the Knowledge Base section and prepare your data sources.

Essential Documents to Include:

• Practice area descriptions and specialties
• Fee structures and consultation policies
• Attorney bios and credentials
• Firm history and core values
• Location and contact information
• Common legal FAQs for your practice areas

Pro Tip: Use the Sitemap option to ingest your entire website at once, or upload individual PDFs for specific practice areas. The AI will automatically break this content into searchable "chunks."

{{blue-cta}}

Step 2: Creating the Current Time Function

Your AI needs temporal context to handle appointment scheduling intelligently. Create a custom function in Voiceflow:

export default async function main(args) {
  // Extract input variables from args
  const {timezone} = args.inputVars;
  
  // Validate that the required input variables are provided
  if (!timezone) {
    return {
      next: { path: 'error' },
      trace: [{ type: "debug", payload: { message: "Please provide a timezone." } }]
    };
  }
  
  try {
    // Get the current time
    const now = new Date();
  
    // Get the current day of the week
    const dayFormatter = new Intl.DateTimeFormat('en-US', {
        weekday: 'long',
        timeZone: timezone
    });
    const dayOfWeek = dayFormatter.format(now);
  
    // Get the current timestamp
    const currentDate = new Date(now.toLocaleString('en-US', { timeZone: timezone })).toISOString();
  
    // Return the extracted variables in outputVars
    return {
      outputVars: {
        current_date: currentDate,
        day_of_week: dayOfWeek,
        timezone: timezone,
      },
      next: { path: "success" },
      trace: [
        {
          type: "debug",
          payload: {
            message: `Successfully retrieved data: current_date=${currentDate}, day_of_week=${dayOfWeek}, timezone=${timezone}`,
          },
        },
      ],
    };
  } catch (error) {
    // Handle errors and log the error message
    return {
      next: { path: "error" },
      trace: [{ type: "debug", payload: { message: "Error: " + error.message } }],
    };
  }
}
  

This function provides your AI agents with real-time context, enabling them to understand references like "tomorrow," "next week," or "Monday afternoon."

Phase 2: Designing the Multi-Agent Conversation System

Step 3: Building "The Referee" - Your Central Routing Agent

The heart of our system is an intelligent routing agent that analyzes caller intent and directs them to the appropriate specialist. Create your first Agent Step with these instructions:

The Referee Agent Instructions:

You are 'The Referee,' the impartial and highly efficient primary AI assistant for Goal Line Legal. Your sole purpose is to intelligently interpret the user's initial request and accurately direct them to the appropriate specialized AI assistant or human team member.

**Communication Style:**
Always communicate as if engaged in a natural, verbal phone conversation. Avoid sounding like a text-based chatbot or using numbered lists. Engage users with a friendly and approachable tone while maintaining clarity and professionalism.

**Your Core Responsibilities:**

**1. Understand User Intent:**
Listen carefully to determine what the user actually needs:
- Starting a new legal case or becoming a client
- Scheduling any type of consultation or appointment  
- General questions about the firm, services, or legal topics
- Speaking directly with a human team member
- Existing client with ongoing case questions

**2. Route Efficiently - Make Quick Decisions:**
Once you understand their intent, immediately direct them:
- **New cases/consultations/scheduling** → Send to "The Coach"
- **General firm info/legal questions** → Send to "The Archivist"
- **Human assistance requests** → Send to "The Dispatcher"
- **Urgent or sensitive matters** → Send to "The Dispatcher"

**3. Stay in Your Lane:**
You are ONLY a router. Do NOT:
- Answer questions about the firm's services or fees
- Provide legal advice or information
- Schedule appointments or collect personal details
- Give detailed explanations about anything
- Try to solve the user's problem yourself

**4. Handle Unclear Requests:**
If you're unsure what they need, ask ONE clarifying question like:
- "Are you looking to start a new legal matter, get information about our services, or speak with someone from our team?"

**5. Keep It Moving:**
Make your routing decision quickly. Don't have long conversations. Your job is to get them to the right specialist as fast as possible.

**Tone & Approach:**
Think of yourself as a professional receptionist who knows exactly where to direct each caller. Be decisive, helpful, and efficient.
  

Configure Three Exit Paths:

Path 1: ToCoachAgent

LLM Description: "Trigger this path when the user expresses an intent related to starting a new legal case, becoming a new client, needing legal representation, discussing a potential new legal matter, or scheduling any form of consultation or 'strategy session'. Look for phrases like 'I need a lawyer,' 'I want to start a case,' 'how do I become a client,' 'book an appointment,' 'schedule a call,' 'consultation,' etc."
  

Path 2: ToArchivistAgent

LLM Description: "Trigger this path when the user asks for general information about Goal Line Legal, its services, practice areas, fees, hours, location, or common legal FAQs. Look for phrases like 'What do you do?', 'Tell me about your firm,' 'How much does it cost?', 'What are your hours?', 'Do you handle sports law?', 'Explain contract disputes.'"
  

Path 3: ToDispatcher

LLM Description: "Trigger this path when the user explicitly asks to speak to a person, lawyer, or team member immediately. Also trigger if the user identifies as an existing client asking about an ongoing case, mentions an urgent legal matter that requires immediate attention, expresses frustration or dissatisfaction with AI assistance, or indicates they have a time-sensitive legal issue that cannot wait. Look for phrases like 'I need to speak to someone now,' 'this is urgent,' 'I'm an existing client,' 'I need a lawyer immediately,' or 'I don't want to talk to a bot.'"
  

Step 4: Creating "The Archivist" - Your Knowledge Expert

The Archivist handles all informational queries using your firm's knowledge base. Position this agent after a KB Search block in your flow.

The Archivist Agent Instructions:

You are 'The Archivist,' the knowledgeable and precise AI assistant for Goal Line Legal. You're speaking with callers on the phone, so keep all responses short, clear, and conversational - like you're having a natural phone conversation.

**Your Responsibilities:**

- **Answer Based on Chunks:** Use only the information from {chunks} to answer their question. Keep responses brief and phone-friendly.

- **Handle No Information:** If {chunks} are empty or insufficient, say something like "I don't have that specific information available right now. Let me connect you with someone who can help."

- **Inform, Don't Advise:** Only share factual information from {chunks}. Never give legal advice or interpret laws.

- **Be Concise:** Give short, clear answers. Remember, they're on the phone - long explanations don't work well.

- **Forward to Appointments:** After answering their question, suggest they speak with The Coach if they want to schedule a consultation about their specific situation.

- **Redirect When Needed:** If they want to start a case or schedule something, use the ReturnToReferee path.

**Phone Conversation Style:**
- Keep responses under 30 seconds when spoken aloud
- Use simple, everyday language
- Pause naturally in your speech
- End with a clear next step

**After answering, always offer:** "Would you like to schedule a consultation to discuss your specific situation? I can connect you with someone who handles appointments."
  

Configure Input Variables:

• originalUserQuestion: Maps to {last_utterance}
• kbRelevantChunks: Maps to {chunks} from KB Search block

Exit Paths:

• EscalateToHuman: For insufficient information or human requests
• ReturnToReferee: When intent changes to booking/new cases
• ScheduleAppointment: When users want to book consultations

Step 5: Building "The Coach" - Your Lead Conversion Engine

The Coach is where prospects become paying clients. This agent handles both new client intake and appointment scheduling.

The Coach Agent Instructions:

You are the dedicated and results-oriented AI assistant for Goal Line Legal. Your sole focus is to efficiently guide prospective clients through the new client intake process and to schedule initial strategy sessions with our legal team. You're speaking with callers on the phone, so keep interactions conversational, supportive, and efficient.

**Current Date Context:**
Today's date is {currentDate} (formatted as YYYY-MM-DD). Use this to understand relative time references like "tomorrow," "next week," or "Monday" when scheduling appointments.

**Your Responsibilities:**
1. **Client Intake:**
   - Use the **initiateNewCaseIntake** function to collect: full name, contact phone, and brief legal challenge summary
   - If missing details, ask directly: "Could I get your full name and best phone number?"

2. **Appointment Scheduling:**
   - Use the **scheduleStrategicHuddle** function to capture preferred date and time
   - When they say "tomorrow," understand that means {currentDate + 1 day}
   - Confirm details: "Perfect! I have you down for [date] at [time]."

3. **Stay Focused:**
   - Only handle intake and scheduling - don't answer general firm questions
   - Keep conversations moving toward completing one of your two functions

4. **After Function Success:**
   - Ask: "Is there anything else I can help you with today?"
   - If they say no or indicate they're done, use the **ConversationComplete** path

**Tone:** Encouraging, precise, and phone-friendly. Keep responses under 30 seconds when spoken.
  

Configure Two Functions:

Function 1: initiateNewCaseIntake

LLM Description: "Use this function when the user clearly expresses an intent to become a new client, start a new legal case, seek legal representation, or asks to discuss a potential new legal matter. This function is designed to systematically gather all essential contact details and a preliminary overview of their legal challenge to create a new lead entry for Goal Line Legal."

Input Variables:
- fullName (String): "The user's complete full name. Essential for client records."
- contactPhone (String): "The user's best contact phone number. Must be a valid phone number format."
- legalChallengeSummary (String): "A brief, high-level description of the user's legal issue or reason for contact."
  

Function 2: scheduleStrategicHuddle

LLM Description: "Use this function when the user explicitly requests to book, arrange, or schedule an appointment, consultation, or a 'strategy session' with a legal professional from Goal Line Legal."

Input Variables:
- preferredDate (String): "The user's desired date for the strategy session. Convert their response to YYYY-MM-DD format using today's date {currentDate} as reference."
- preferredTime (String): "The user's preferred time for the strategy session, formatted in 24-hour format (HH:MM). Convert their response: '10 AM' becomes '10:00', '2 PM' becomes '14:00'."
  

Exit Paths:

• ConversationComplete: After successful function completion
• EscalateToHuman: For complex issues or user frustration
• ReturnToReferee: When intent changes outside scope

Step 6: Creating "The Dispatcher" - Your Human Handoff Specialist

The Dispatcher ensures seamless escalation to human team members when AI reaches its limits.

The Dispatcher Agent Instructions:

You are 'The Dispatcher,' the crucial AI assistant for Goal Line Legal, specializing in preparing and coordinating seamless human hand-offs. Your core mission is to ensure all essential information is confirmed and accurately packaged for our legal team before notifying them about a required callback or follow-up.

**Your Responsibilities:**

**Confirm/Gather Contact:** Verify or politely collect the user's full name and best contact phone number.

**Clarify Query:** Confirm the main reason or query for the hand-off to provide the human team with maximum context.

**Trigger Notification:** Once all necessary information is confirmed, use the notifyColleagueOfHandoff function to send an email summary to the relevant legal professional.

**Manage Expectations:** Inform the user clearly about the next steps (e.g., when they can expect a callback).

**Conclude Professionally:** End the conversation gracefully after the hand-off is coordinated.

**Voice Style:** Always communicate as if engaged in a natural, verbal phone conversation; avoid sounding like a text-based chatbot or using numbered lists.

**Important:** You must not attempt to answer legal questions or provide advice. Your sole purpose is to facilitate the hand-off process with complete and accurate information.
  

Configure Function: notifyColleagueOfHandoff

LLM Description: "Use this function only when the user has confirmed all their contact details (full name and phone number) and the reason for the human hand-off, and they are ready for the legal team to be notified for a callback."

Input Variables:
- handoffFullName (String): "The user's complete full name for the legal team to address correctly."
- handoffPhoneNumber (String): "The user's best contact phone number for callback."
- handoffQuerySummary (String): "A concise summary of the user's legal query or reason they require human assistance."
  

Phase 3: Building Your Make.com Business Automation

Now we'll create the three Make.com scenarios that power your backend business processes.

Step 7: Setting Up Google Sheets CRM Database

Create a Google Sheet named "Legal Answering Service" with three tabs:

Tab 1: initiateNewCaseIntake

• Column A: Full Name
• Column B: Phone
• Column C: Legal Challenge Summary

Tab 2: scheduleStrategicHuddle

• Column A: Start time
• Column B: Date

Tab 3: notifyColleagueOfHandoff

• Column A: Full name
• Column B: Phone number

{{blue-cta}}

Step 8: Create New Client Intake Automation

Import the initiateNewCaseIntake.blueprint.json into Make.com or build manually:

Workflow Structure:

1. Custom Webhook: Receives data from Voiceflow Coach agent
2. Google Sheets Node: Adds new row with prospect information
3. Email Node: Sends professional lead notification

Key Configuration:

• Webhook receives: fullname, contactPhone, legalChallengeSummary
• Email template uses professional HTML formatting with firm branding
• Subject line: "New Intake" for immediate team attention

Step 9: Create Appointment Booking Automation

Import the scheduleStrategicHuddle.blueprint.json:

Workflow Structure:

1. Custom Webhook: Receives scheduling data from Voiceflow
2. Google Sheets Node: Logs appointment request details
3. Email Node: Notifies team of new consultation booking

Professional Email Template:

<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>New Strategy Session Booked: Goal Line Legal</title>
</head>
<body style="font-family: Arial, sans-serif; line-height: 1.6; color: #333;">
  <div style="max-width: 600px; margin: 20px auto; padding: 20px; border: 1px solid #ddd; border-radius: 8px;">
    <h2 style="color: #2196F3;">New Strategy Session Request</h2>
    <p>A user has requested to schedule an initial strategy session via the Goal Line Legal AI Assistant.</p>
    <table width="100%" cellpadding="0" cellspacing="0" border="0" style="margin-top: 20px;">
      <tr>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee; width: 30%; font-weight: bold;">Preferred Date:</td>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee;">{{preferredDate}}</td>
      </tr>
      <tr>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee; width: 30%; font-weight: bold;">Preferred Time:</td>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee;">{{preferredTime}}</td>
      </tr>
    </table>
    <p style="margin-top: 20px; font-size: 0.9em; color: #666;">This notification was automatically generated by The Coach AI Assistant.</p>
  </div>
</body>
</html>
  

Step 10: Create Human Handoff Automation

Import the notifyColleagueOfHandoff.blueprint.json:

Workflow Structure:

1. Custom Webhook: Receives urgent handoff requests
2. Google Sheets Node: Logs handoff details for follow-up tracking
3. Email Node: Sends high-priority notification to legal team

Urgent Email Template:

<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>Urgent: Human Handoff Request - Goal Line Legal</title>
</head>
<body style="font-family: Arial, sans-serif; line-height: 1.6; color: #333;">
  <div style="max-width: 600px; margin: 20px auto; padding: 20px; border: 1px solid #ddd; border-radius: 8px; background-color: #FFEBEE;">
    <h2 style="color: #F44336;">Urgent: Human Assistance Required!</h2>
    <p>A user requires direct human assistance from Goal Line Legal. Please follow up with them as soon as possible.</p>
    <table width="100%" cellpadding="0" cellspacing="0" border="0" style="margin-top: 20px;">
      <tr>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee; width: 30%; font-weight: bold;">Full Name:</td>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee;">{{handoffFullName}}</td>
      </tr>
      <tr>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee; width: 30%; font-weight: bold;">Phone Number:</td>
        <td style="padding: 8px 0; border-bottom: 1px solid #eee;">{{handoffPhoneNumber}}</td>
      </tr>
    </table>
    <p style="margin-top: 20px; font-size: 0.9em; color: #666;">This urgent notification was automatically generated by The Dispatcher AI Assistant.</p>
  </div>
</body>
</html>
  

Phase 4: Connecting Your Complete System

Step 11: Configure Voiceflow API Integrations

Connect each agent's functions to your Make.com webhooks:

In The Coach Agent:

• Function 1: initiateNewCaseIntake → Make.com webhook URL
• Function 2: scheduleStrategicHuddle → Make.com webhook URL

In The Dispatcher Agent:

• Function: notifyColleagueOfHandoff → Make.com webhook URL

API Configuration:

• Method: POST
• Content-Type: application/json
• Body: Include all function input variables as JSON

Step 12: Complete Flow Testing

Test your entire system with this sequence:

1. Start Block Test: Verify welcome message and variable capture
2. Referee Routing: Test all three routing paths with different inputs
3. Archivist Knowledge: Ask questions that should find KB answers
4. Coach Functions: Test both intake and scheduling workflows
5. Dispatcher Handoff: Verify urgent escalation process
6. Integration Validation: Confirm Google Sheets data capture and email delivery

Sample Test Conversations:

For Referee → Coach → Intake:

• Input: "I need legal representation for a contract dispute"
• Expected: Routes to Coach → Collects name, phone, case summary → Emails team

For Referee → Archivist → Knowledge:

• Input: "What practice areas do you handle?"
• Expected: Routes to KB Search → Returns relevant info → Offers consultation

For Referee → Dispatcher → Handoff:

• Input: "This is urgent, I need to speak to a lawyer immediately"
• Expected: Routes to Dispatcher → Collects contact info → Sends urgent email

Phase 5: Advanced Optimizations & Scaling

Step 13: Phone System Integration

For true 24/7 operation, integrate with phone systems:

VoIP Integration Options:

• Twilio Voice API for inbound call handling
• 3CX or FreePBX for existing phone system integration
• Voiceflow's native phone channel deployment

Configuration Requirements:

• Dedicated phone number for AI assistant
• Voicemail fallback for system maintenance
• Call recording for quality assurance and training

Step 14: Advanced CRM Integration

Enhance your system with professional CRM connections:

Supported Integrations:

• HubSpot for complete sales pipeline management
• Pipedrive for deal tracking and follow-up sequences
• Salesforce for enterprise-level client management
• Clio or MyCase for legal practice management

Integration Benefits:

• Automatic lead scoring and qualification
• Appointment scheduling with conflict detection
• Follow-up sequence automation
• ROI tracking and conversion analytics

Step 15: Multi-Language Support

Expand your reach with international capabilities:

Implementation Strategy:

• Duplicate agent workflows for each target language
• Use Make.com's translation modules for dynamic content
• Create language-specific knowledge bases
• Configure routing based on caller language preference

Real-World Results: What to Expect

Based on implementations across various law firms, here's what you can typically expect:

Week 1-2: 60-80% of after-hours contacts engage with the AI assistant Month 1: 40-55% conversion rate from conversation to qualified appointment
Month 3: 80-90% reduction in missed opportunities outside business hours Ongoing: 24/7 professional client experience with complete business integration

One personal injury firm saw a 400% increase in consultation bookings within the first month, simply because prospects could schedule meetings instantly during evening hours when they typically research legal options.

Download Ready-to-Use Templates

Want to skip the setup and start with working templates? I've created a complete package:

Voiceflow Template:

• Pre-configured 4-agent system with all prompts and paths
• Complete conversation flows with error handling
• API integration blocks ready to connect
• Sports law firm theme with customizable branding

Download the template

Make.com Automation Templates:

• New client intake and email notification workflow
• Appointment booking and calendar integration
• Human handoff and urgent notification system
• Google Sheets CRM structure and formulas

Download the Make.com blueprints

Advanced Customizations

Once your basic AI receptionist is running, consider these enhancements:

Smart Scheduling Integration: Connect to Google Calendar or Calendly for real-time availability checking and automatic appointment confirmation.

Multi-Practice Area Routing: Create specialized agents for different legal specialties (personal injury, corporate law, family law) with practice-specific knowledge bases.

Client Portal Integration: Connect to existing client portals for existing client authentication and case status updates.

Advanced Analytics: Track conversion rates, popular inquiry types, and optimal response patterns through detailed conversation analytics.

Follow-Up Automation: Set up automatic reminder sequences for scheduled consultations and follow-up communications for prospects who don't immediately book.

Troubleshooting Common Issues

Problem: "AI gives inconsistent responses to similar questions" Solution: Refine your agent instructions with more specific examples and edge case handling. Use the conversation logs to identify patterns and improve prompts.

Problem: "Make.com webhooks not receiving data"
Solution: Verify your webhook URLs are correctly configured in Voiceflow functions. Test the connection using Make.com's webhook testing tools.

Problem: "Knowledge base returns irrelevant information" Solution: Review your knowledge base content for duplicate or contradictory information. Use more specific document titles and clear section headers.

Problem: "Appointments being scheduled outside business hours" Solution: Add business hour validation to your Coach agent instructions and include available time slots in the scheduling function.

Problem: "Email notifications not being delivered" Solution: Check your Gmail API authentication and ensure sending limits aren't exceeded. Consider using a dedicated business email service for high-volume notifications.

Scaling Your AI Assistant Business

Your law firm AI assistant is just the beginning. Here's how to evolve it into a competitive advantage:

Template Development: Create industry-specific versions for medical practices, real estate agencies, consulting firms, and other professional services.

White-Label Solutions: Offer your AI assistant system to other law firms as a managed service, creating a new revenue stream.

Integration Marketplace: Develop connections to popular legal software like Clio, LawPay, and DocuSign for comprehensive practice management.

Training and Certification: Create educational content teaching other professionals how to implement and customize their own AI receptionist systems.

Performance Analytics: Build detailed reporting dashboards showing ROI, conversion rates, and optimization opportunities for continuous improvement.

Your AI-Powered Legal Practice

Building a Voiceflow AI assistant integrated with Make.com and Google Workspace isn't just about automation—it's about creating intelligent client experiences that scale your practice and eliminate missed opportunities. You've just built a system that combines natural conversation with sophisticated business process automation.

The combination of Voiceflow's multi-agent capabilities, Make.com's integration power, and professional email systems creates something greater than the sum of its parts—a truly intelligent business assistant that works around the clock, captures qualified leads, and provides instant value to potential clients.

Ready to implement this for your law firm? Start with the templates above, or reach out to me (Abdullah) if you need help customizing this system for your specific practice area and business requirements. As an AI automation specialist, I help law firms implement these exact workflows with proper customization and optimization tailored to their unique client needs and practice management systems.

The future of legal client acquisition is here—and it never sleeps, never misses a call, and never lets a qualified prospect slip away. Your 24/7 AI legal assistant is ready to transform how you capture and convert clients.