AI Text Agents - Overview

Overview

AI Text Agents are the backbone of automated conversations in brayv.ai. They use advanced AI to have natural, intelligent text message conversations with your contacts - answering questions, gathering information, booking appointments, and moving leads through your sales process without manual intervention.

What AI Text Agents Do

An AI Text Agent can:

• Respond to text messages in natural, conversational language
• Gather information from contacts and save it to their profile
• Book appointments by checking calendar availability and scheduling
• Follow up with contacts automatically on a schedule you define
• Take actions based on conversation context (add to journeys, send notifications, end journeys, etc.)
• Transfer conversations by turning off AI so you can jump in manually
• Answer questions using your knowledge base and business context

Two Types of AI Text Agents

Global Agents

Purpose: Customer service and triage for contacts not in an active journey.

How they work:

• Answer contacts who text in but are not currently in an active journey
• Do not proactively follow up with leads
• Serve as a safety net for "stray" contacts:
  • Contacts texting back after their journey ended
  • Cold texts from people who found your number
  • Chat widget conversations
  • General customer service questions

Best Practice - Handoff to Journeys: Global agents should be configured with conditional tools to add contacts to appropriate journeys when they express interest. For example:

• Trigger: "the user indicates they are interested in a new pool"
• Action: Add to Journey: "New Pool Build Leads"

Once a contact is added to a journey, the journey's agent at the Starting Line stage takes over the conversation.

Default Global Agent: You can set one global agent as your default (Settings > Text AI). This agent handles any contact with text AI enabled but no specific agent assigned.

Where to create/edit: AI Agents page > Text Agents tab > New Agent (or edit existing)

Journey Agents

Purpose: Engage, qualify, follow up with leads, and achieve specific goals within a customer journey.

How they work:

• Assigned to specific stages within customer journeys
• Proactively follow up with contacts to achieve the stage goal
• Take over when a contact enters their stage
• Can only be created/edited from within the journey stage they belong to (not from the AI Agents page)

Journey Agent at Start Stage: This is your primary lead engagement agent. Configure it with:

• Information gathering: What details do you need? (service description, address, timeline, etc.)
• Booking Options: Calendar selection so AI can schedule appointments
• Or Completion Goal For non-appointment goals (eg lead qualification, where you want to verify certain information, but not book an appointment)
• Conditional tools: Take actions on specific things that happen in the conversation (beyond just responding to the contact), ie, what to do when the contact says they aren't interested, says they want pricing, or asks not to be contacted again.
• Follow-up sequences: Automated follow-ups to move leads toward booking or qualification (completion goal)

Journey Agent at Appointment Scheduled Stage: After the goal (appointment) is achieved, this agent:

• Answers questions leading up to the appointment
• Provides reschedule and cancellation links (if you give it access)
• Does NOT need follow-up sequences configured (goal already achieved)
• Handles pre-appointment concerns

Journey Agents at Later Stages (Advanced): For more complex journeys, you might have agents at post-goal stages:

• Asking for feedback after service completion
• Following up on estimates
• Requesting reviews
• Upselling additional services

Where to create/edit: Within the specific journey stage (Journeys > [Journey Name] > [Stage Name] > Configure Agent)

Where to view: AI Agents page > Journey Agents tab (view only, not editable here)

The Typical 3-Agent Setup

Most accounts need just 3 AI Text Agents:

1. Default Global Agent

• Role: Customer service and journey handoff
• Scope: Global
• Configuration:
  • General customer service instructions
  • Conditional tools to add contacts to journeys based on expressed interest
  • No follow-up sequences
• Example prompt focus: "Answer questions about our services. If they express interest in [service], add them to the appropriate journey."

2. Journey Start Stage Agent

• Role: Lead engagement, qualification, and appointment booking
• Scope: Journey-specific (Start stage)
• Configuration:
  • Information gathering instructions
  • Calendar booking enabled
  • Conditional tools for "not interested" and "do not contact" scenarios
  • Follow-up sequences to re-engage unresponsive leads
• Example prompt focus: "Qualify the lead, gather service details and address, book an appointment."

3. Journey Appointment Scheduled Agent

• Role: Pre-appointment support
• Scope: Journey-specific (Appointment Scheduled stage)
• Configuration:
  • Instructions about appointment details and rescheduling
  • No follow-up sequences (goal already achieved)
  • Conditional tools for rescheduling/cancellation alerts
• Example prompt focus: "Answer questions about the upcoming appointment, provide reschedule/cancel links if needed."

This minimal setup covers 90% of use cases while keeping your system simple and maintainable.

How Contacts Flow Through the System

The Ideal Flow (Lead Sources → Journey)

Lead Source (Form/Webhook/Facebook Lead Form)
    ↓
Automatically added to Journey
    ↓
Start Stage Agent takes over
    ↓
Engages, gathers info, follows up
    ↓
Books appointment
    ↓
Moves to Appointment Scheduled stage
    ↓
Appointment Stage Agent takes over
    ↓
Does not proactively reach out to contact, but answers questions until appointment occurs
    ↓
Journey ends after appointment (or moves to later stages)

Global Agent Flow (Strays & Customer Service)

Contact texts in (not in active journey)
    ↓
Default Global Agent responds
    ↓
Option A: Contact expresses interest → Added to Journey for relevant product/service → Journey agent takes over
Option B: Contact just has questions → Global agent answers (maybe sends particular team members notifications if contact makes particular requests) → Conversation ends

Creating and Editing AI Text Agents

All agent creation and editing happens through the AI Agent Editor Modal. (Dialog that opens up on top of the page you are on and has all agent configuration options)

Creating a Global Agent

Step 1: Navigate to AI Agents page

Step 2: Click "New Agent" button in the Text Agents tab

Step 3: The AI Agent Editor Modal opens with these tabs:

• Options: Basic settings (name, model, temperature, response delay)
• Instructions: System prompt (what the agent does and how it behaves in the conversation)
• Actions: Conditional tools (trigger actions based on conversation)
• Field Entry: Field extractors (save conversation data to contact fields)
• Follow-ups: Follow-up sequences (automated follow-up messages - easy to configure just click generate with AI to get a full sequence which you can then adjust as you like)
• Booking/Goal: Calendar booking configuration, or Completion goal configuration (conversations with a non-appointment goal)
• Audits: Version history and changes

Step 4: Configure each section (see detailed sections below)

Step 4 Easy Mode: Click the "Edit with AI" button. Tell Ava (our AI assistant) what you want. Ava configures instructions, conditional tools, and field entry. You can then click apply and review before saving.

Step 5: Click Save

Creating a Journey Agent

Step 1: Navigate to Journeys page

Step 2: Open the journey you want to add an agent to

Step 3: Click on the specific stage where you want the agent

Step 4: Click "Add Text Agent" button or "Edit Agent"

Step 5: The AI Agent Editor Modal opens (same interface as global agents)

Step 6: Configure and save

Important: Journey agents can only be edited from within their journey stage, not from the AI Agents page.

AI Agent Editor Modal Overview

Options Tab

Basic configuration settings:

• Name: Descriptive name for the agent (e.g., "Global Customer Service", "Pool Leads - Start Stage")
• Description: Internal notes about what this agent does
• Model: AI model to use (gpt-5 recommended for most cases)
• Temperature: Creativity level (0.7 is default, lower = more consistent, higher = more creative)
• Max Tokens: Maximum response length (500 default - that's huge, you can significantly lower if you are experiencing more wordy responses than you would like/if agent ignores character limit in Instructions)
• Response Delay: Seconds to wait before responding (40 default, makes AI seem more human, allows AI to respond to double texts with just one message)
• Enable Smart Delay: AI decides response timing based on message complexity

Instructions Tab

System prompt - the agent's brain:

This is where you define:

• Role: Who is the agent? ("Your name is Alex. You are a customer service rep for {account.name}")
• Goal: What should the agent accomplish?
• Rules: Constraints and guidelines (character limits, emoji usage, pricing policies, etc.)
• Context: Business information the agent should know

You can use variables in your instructions:

• {contact.first_name} - Contact's first name
• {contact.hook} - What they're interested in
• {contact.friendly_source} - How they found you
• {account.name} - Your business name
• {account.company_context} - Your business description
• And many more (see variable picker in editor)

AI Editor Assistant: Click the AI assistant icon to use AI to help write or modify your instructions, conditional tools and their actions, and field extractors. You can describe what you want, and the AI will update the prompt for you. (HIGHLY RECOMMENDED. Ava will get you 95-100% of the way to what you want in an agent in a couple minutes).

Actions Tab (Conditional Tools)

Trigger actions based on conversation context.

Conditional tools watch the conversation and automatically trigger actions when certain conditions are met.

Quick Add Templates: Common pre-built tools you can add with one click:

• Not Interested: Ends journey, turns off AI
• Stop Contact Ends journey, marks contact Do Not Text, turns off AI
• Bad Contact Info: Marks DND, ends journey, turns off AI
• Zip Code Check: Evaluates if contact is in service area based on zip code, takes action accordingly. (you must add in your list of zip codes to fully configure)

Custom Tools: Create your own with:

• Name: Internal reference
• Trigger Conditions: Plain English description of when to activate - what the AI uses to decide whether to trigger the tool (e.g., "the user indicates they want pricing for a new HVAC system")
• Actions: What to do when triggered (see all available actions in detailed doc)

See: Conditional Tools Guide for detailed information

Field Entry Tab (Field Extractors)

Automatically save information from conversations to contact fields.

Field extractors "listen" to conversations and extract information to save to contact records.

Example:

• Field: street_address
• Description: "the user's street address. Ex: '123 Main St'"
• Allow Overwrite: Whether to update if field already has a value

When the contact mentions their address in conversation, it's automatically saved to their contact record.

See: Field Extractors Guide for detailed information

Follow-ups Tab

Automated follow-up message sequences.

Configure a series of messages to send if the contact doesn't respond.

Settings:

• Enable Follow-ups: Turn the sequence on/off
• Pause on Reply: Hours to wait after contact responds before resuming sequence
• Time Gate: Only send during specific hours/days (recommended, only timegates the proactive followup, Agent still responds 24/7 when contact sends you a message)
• Messages: Sequence of messages with delays between each

Important:

• Only use follow-ups on Journey Start Stage agents (to push toward goal)
• Do NOT use on Appointment Scheduled stage agents (goal already achieved, if you want appointment reminders, configure in Calendar settings)

See: Follow-up Sequences Guide for detailed information

Booking/Goal Tab

Three options:

1. "Neither" (for agents that are not proactively pursuing a goal)
2. "Calendar Booking" (for agents whose end goal is booking an appointment)
3. "Completion Goal" (for agents whose goal is to gather/verify certain information but not book an appointment)

Configure calendar appointment booking.

Modes available:

1. No Booking: Agent cannot book appointments

2. Account Default: Agent books in your account default calendar (can be set in Settings > Business > Account Defaults)

2. Specific Calendar: Agent books to one designated calendar

• Select the calendar you want the agent to book into
• AI checks availability and books appointments
• Simplest option

3. AI Selection: Agent intelligently chooses which calendar/user to book with

• Provide instructions for selection logic based on contact data (e.g., "if contact zip code is 11706, 11707, 11708... book into Calendar A, otherwise, book into calendar B")
• AI follows your rules to select appropriate calendar
• Most flexible option
• Use the AI chat assistant to help turn your instructions into a quick script (code) the AI can run instantly to access the correct calendar.
• Advanced: include user selection (which team member to book with) as well based on contact data or user appointment counts.

See: Booking Configuration Guide for detailed information

Configure completion goal

Only available if Calendar booking is not selected. (Agents with calendar booking goals can, and often do, gather information before attempting to book)

• Useful for when you want to gather/verify certain info/"qualify" your lead (but not book an appointment)

Select Target Stage

• This is the stage the Agent will move the contact's opportunity to once it has reached its goal
• Stage must be either a primary or secondary "goal" stage
• Stage must be in the same Journey the agent is in
• Stage must be later in the Journey then the stage the agent you are editing is assigned to
• (only eligible stages will show)

Configure 'Completion Criteria'

• Describe what constitutes reaching the goal
• Agent evaluates this criteria and determines if and when to mark the conversation as successfuly completed (move opportunity to designated target stage)

Configure Actions

• None required
• Optionally send emails, sms, or in app notifications to specific members of your team when the goal is achieved
• Even if you don't configure actions, the basic functionality will still work: moving the contact's opportunity to the designated target stage show journey shows contact as reaching the goal.

Audits Tab

Version history and change tracking.

View:

• All changes made to the agent
• Who made each change
• When changes were made
• Ability to revert to previous versions (through version selector at the top of the interface - 2 most recent versions and a pinned version)

Agent Configuration Variables

Throughout your agent configuration, you can use variables that get replaced with actual data:

To access variables, just hit the curly bracket key on your keyboard.

Contact Variables

• {contact.first_name}, {contact.last_name}, {contact.full_name}
• {contact.email}, {contact.phone}
• {contact.business_name}
• {contact.street_address}, {contact.city}, {contact.state}, {contact.zip_code}
• {contact.hook} - What they're interested in
• {contact.friendly_source} - How they found you
• {contact.service_description} - Details about their request
• Custom fields: {contact.custom_field_name}

Account Variables

• {account.name} - Your business name
• {account.company_context} - Your business description

Appointment Variables (for appointment stage agents)

• {appointment.details_and_reschedule_cancel} - Full appointment details with links

Calendar Variables

• Available calendars and users (in booking configuration)

Best Practices

Writing Effective Instructions

Be Specific:

• ❌ "Be helpful"
• ✅ "Answer questions about our HVAC services. If asked about pricing, say 'I can have a specialist confirm pricing details for you.'"

Use the Slot-Filling Pattern (Start Stage Agents): Structure your prompt to gather information step-by-step:

## Conversation Goal
Work through the following goals in order. Treat each goal as a "slot" with a current value.

1) **Service Description**
- current_value: {contact.service_description}
- policy: always_ask
- example: "Can you tell me a little bit about the project you have in mind?"

2) **Address**
- current_value: {contact.street_address}, {contact.city}, {contact.zip_code}
- policy: always_ask
- example if present: "I have your adress as {contact.street_address}, {contact.city} {contact.zip_code}, is that correct?"
- example if missing: "Can you provide your city, state, and zip code?"

3) **Book Appointment**
- Tell the user: "Great, sounds like we can help. Let's find a time that works for you. I see these times available..."

Keep Responses Short:

• Set character limits (250-300 characters is a good default)
• People prefer short texts over long paragraphs

No Emojis (Usually):

• Avoiding emojis in AI responses might fit your brand
• Big benefit is it significantly reduces the number of SMS "segments" you send (SMS with emojis are encoded differently and segment size reduces from 160 characters to ~ 70)

Conditional Tools Strategy

Always Include These for Start Stage Agents:

1. Not Interested - Gracefully end journey, turn off AI, notify team if desired
2. Bad Contact Info - Mark DND, end journey
3. Add to Journey (for global agents or if you have multiple journeys for different customer interests) - Hand off to appropriate journey

Use Rate Limiting:

• Set rate_limit_minutes on notification actions to avoid getting notified about the same thing multiple times
• Example: Only notify team once per 15 minutes about a "call back request" response (contact may say, "i want a call back" - tool gets triggered. Agent responds "no problem, I'll make sure someone contacts you", Contact responds again, "great, yeah I really need someone to call" - rate limit prevents the tool from triggering again, because this is really the same request)

Follow-up Sequences

Do Use For:

• Journey start stage agents (pushing toward booking)
• Re-engaging cold leads
• Moving stalled conversations forward
• When the agent is actively pursuing a goal (you want the contact to take a specific next step, like provide certain info or schedule an appointment)

Don't Use For:

• Appointment scheduled stage agents (goal already achieved)
• Global agents (they're reactive, not proactive)
• Post-goal stages where you're not actively trying to get the customer to take a next step (eg, maybe your team is preparing a proposal)

Follow-up Message Tips:

• Start friendly and helpful
• Reference their original interest: "Hi {contact.first_name}, following up about your {contact.hook} request"
• Provide value in each message or keep it super simple "Is now a better time to chat?"
• Know when to stop (recommendation: new leads 10 attempts max over 2 months, old leads for sale/promotion reactivation 2-3 attempts, over short time frame)

Booking Configuration

For Most Businesses: Use Specific Calendar mode if:

• Appointments are split up relatively evenly between reps, there are no specific territories per rep or specific reps that handle specific appointment types
• Calendar selection is straightforward
• You want simplicity

Use AI Selection mode if:

• Multiple technicians/staff with different specialties
• Territory-based assignment needed
• Complex selection logic based on different contact data
• Appointment count preferences (i.e. George should get 3 appointments for every 2 that John gets)

AI Selection Instructions: Be clear and specific:

• Add in the specific variables the AI should consider when booking an appointment. Add the specific calendars/users as well. (use the curly bracket key to open variable selector)
• "If the {contact.service_type} is "plumbing", book in Plumbing Calendar. If "HVAC", book in Air Calendar."
• Always provide a fallback option: "otherwise, book in Calendar Z"

Setting Your Default Global Agent

Your default global agent handles any contact with text AI enabled but no specific agent assigned.

To set it:

1. Go to Settings > Text AI
2. Select your global agent from the Default Agent dropdown
3. Configure other settings:
   • Enable AI for new contacts: Auto-enable text AI when contacts are created (recommended)
   • Excluded Tags: Contacts with these tags won't receive AI responses regardless of whether the AI is on or off for them (e.g., "cranky customer", "vip-manual-only")
4. Save

Using the AI Editor Assistant

The AI Editor Assistant is a chat interface within the agent editor that can help you:

Configure Instructions:

• "Add a rule that the agent should never quote pricing"
• "Update the goal to focus on booking appointments, not just answering questions"
• "Make the tone more casual and friendly"

Create Conditional Tools:

• "Add a tool that sends me a notification when someone asks about our premium services"
• "Create a tool to end the journey if they say they're not interested"

Configure Field Extractors:

• "Extract the customer's phone number, email, and address from conversations"

What it won't do:

• Build your follow up sequence - just click on that tab, there's an easy generate with AI option there.
• Configure booking options - either select a specific calendar, or use AI to help configure your AI selection instructions in the BOOKING/GOAL tab.
• Configure your completion goal - it can suggest what to write for the completion criteria, but you must click the BOOKING/GOAL tab to configure a completion goal.

Access the AI Assistant: Click the sparkle icon or "AI Assistant" button in the agent editor modal.

Monitoring Agent Performance

Conversations

View all ongoing conversations on the conversations screen

• Jump into any conversation you want
• Turn off AI/Turn back on AI as needed with simple toggle
• Update Journey opportunities (mark as "lost" or "won") to end journeys and followups when appropriate
• Book appointments directly to move contact along in journey

AI Text Agent Logs

All logs are a pair of contact message and the response the AI gave. View all AI text agent responses and tool execution:

Location: AI Agents page > Logs tab

What you can see:

• Sort by Agents and message status
• Search by contact message, agent response, contact id, specific errors etc.
• Every response sent by AI agents (sequenced follow ups do not show up here)
• Which agent handled the conversation
• Tokens used (cost tracking)
• Response times
• Errors or issues
• Conditional tools that triggered
• Reasoning (why the AI used a conditional tool)
• Field extractors that fired

Use logs to:

• Troubleshoot issues
• Improve agent instructions
• Monitor AI performance
• Track costs
• Audit conversations

Testing Agents

Before deploying an agent:

1. Use the Test button on the agent card (AI Agents page, or on the Agent Card within a Journey Stage)
2. Have a full conversation with the agent
3. Verify it responds appropriately
4. Check that conditional tools trigger correctly
5. Optional - prefill data for test contact so you can see how the agent behaves when certain contact info is already present. Just click the Pre-Fill Fields button at the top of the testing interface and fill fields like name, zip code, hook, etc.

Notes about Testing

• Conditional Tools still trigger so you can ensure your trigger conditions work the way you want, but their configured actions don't actually do anything (you wont get an email notification etc, because its not attached to real contact)
• Field Extractors still fire, and show you the value they would have saved to a contact, but again this does not save anywhere (because there is no real contact associated with the conversation)
• Booking does NOT use your calendar configuration from the agent, it uses a dummy test calendar so you don't clog up your calendar with fake appointments. (still must have booking enabled in the agent settings to test a booking interaction)

Common Use Cases

Use Case 1: Home Services Company

Setup:

• Global Agent: Answers customer service questions, adds new inquiries to journey
• Start Stage Agent: Qualifies leads, gathers service details and address, checks territory, books consultation
• Appointment Stage Agent: Answers pre-appointment questions, provides reschedule/cancel links

Use Case 2: Professional Services (Law, Accounting, etc.)

Setup:

• Global Agent: Initial intake, adds potential clients to consultation journey
• Start Stage Agent: Gathers case details, schedules consultation calls
• Post-Consultation Agent: Follows up on proposal, answers questions about next steps

Use Case 3: E-commerce/Product Business

Setup:

• Global Agent: Product questions, order status, returns - with handoff to journeys for new sales opportunities
• Sales Journey Agent: Qualifies interest, gathers requirements, books demo/consultation

Tips & Best Practices

Start Simple

Begin with the 3-agent setup. Don't over-engineer.

Use Default Templates

The default prompts provided by brayv.ai are solid starting points. Customize from there.

Use Ava

The AI editor assistant interface is very powerful and can help you get where you want to go with your agents much faster then manually editing (typically). But always feel free to manually edit anything as well. Always review what Ava does before saving.

Test Thoroughly

Test your agents before going live. Have real conversations to see how they handle edge cases.

Iterate Based on Logs

Review your agent logs weekly. Look for:

• Misunderstood prompts
• Questions the agent couldn't answer (add to knowledge base)
• Patterns in how people respond
• Opportunities to improve instructions

Leverage Variables

Use contact variables ({contact.first_name}, {contact.hook}, etc.) to make conversations feel personal.

Keep Instructions Updated

When your business changes (new services, pricing policies, procedures), update your agent instructions.

Monitor Token Usage

AI costs money per token used. Monitor your logs to:

• Keep response lengths reasonable (max_tokens setting if needed, typically rule in instructions about character length is fine)
• Avoid unnecessary back-and-forth
• Optimize prompts for efficiency, Don't over explain to your agent.

Troubleshooting

Problem: Agent not responding to contacts

Solution:

• Verify contact has text AI enabled (Contact detail page > AI Text Agent toggle)
• Check contact doesn't have "Do Not Text" enabled
• Verify agent is active
• Check your account has an active phone number configured
• Review AI Text Agent Logs for errors

Problem: Agent giving wrong answers

Solution:

• Review and refine the system prompt (Instructions tab)
• Add specific rules about things it should not say
• Add relevant information to your knowledge base
• Check if the agent if your business context is up to date: Settings > Business > Account Defaults

Problem: Conditional tools not triggering

Solution:

• Make trigger conditions more specific and clear
• Test in the agent test interface
• Review logs to see what the AI is detecting
• Ensure the tool is enabled

Problem: Agent booking appointments incorrectly

Solution:

• Review booking mode configuration
• Ensure a booking mode is selected and configured. If none is selected or the selected mode is not fully configured, AI will not be able to book.
• If using AI Selection mode, check logs to see what calendar/user it chose and why
• Check calendar availability settings
• Us the test functionality in AI Selection mode (Booking/Goal tab > Calendar Booking) to test if is selecting the calendar/user you want.
• Verify calendar has users in it. If specifying specific users in AI selection mode, ensure those specific users are selected in the calendars you have told the AI to use.

Problem: Follow-ups not sending

Solution:

• Verify follow-up sequence is enabled
• Check contact (below assigned AI agent will display next scheduled follow up)
• Ensure contact hasn't responded (pauses sequence, will resume)
• Check that contact still has text AI enabled
• Review contact's journey status (might have ended)
• Review contact's DND status

Problem: Agent responses too slow

Solution:

• Reduce response_delay_seconds in Options tab
• Consider using a different AI model in Options tab
• Slow sometimes is better than fast. (eg, delaying 40 seconds helps the AI not be annoying. User texts in "my address is 123 Main St" then ten seconds later "Wilsonville, AL 35186" - its better for the AI to be "slow" and respond to this all as one message, instead of being "fast" and saying "thanks for providing your street address, can you also provide your city and zip?" right when the user is sending it. "Slow" sometimes means more human.)
• Sometimes AI model providers, like OpenAI go "down" for extended periods of time. Not to worry, brayv.ai is built to be resilient, with multiple layers of fallbacks, but when events like this occur response times will be slower as the system cylces through different layers of fallbacks to ensure your contact gets a response.
• If you experience the AI taking more than 5 minutes to respond, and you confirm the AI was on at the time of the contact's message, please reach out to support so we can help!

Problem: Too many notifications from conditional tools

Solution:

• Add rate_limit_minutes to notification actions
• Be more specific with trigger conditions to reduce false positives
• Consolidate multiple notifications into one

Related Features

• Customer Journeys - Orchestrate contact flow and agent assignment
• Conditional Tools - Detailed guide to creating and managing conditional tools
• Field Extractors - Extract and save conversation data automatically
• Follow-up Sequences - Configure automated follow-up messages
• Booking Configuration - Set up calendar appointment booking