🤖 AI Voice & Chat5.1GHL

A.I. Chat | Messages

A multi-branch AI engine that answers inbound chat across every channel via webhook responses and escalates to a human when the bot can't reply.

3 triggers 18 steps 3 branches Advanced

What it does

This is one of the most architecturally sophisticated automations in the entire system — a wide, multi-branch AI conversation workflow that manages inbound chat messages across every active channel simultaneously. The automation detects the message source, identifies the contact, and routes the conversation into the appropriate AI response branch. Across eight or more parallel paths it handles different conversation intents — enquiries, objections, booking requests, existing client messages, and unrecognised inputs — each triggering its own AI-generated response sequence. Deeper in the workflow, AI message analysis branches determine whether the conversation is active, stalled, or complete, and route accordingly. The final layer handles escalation logic — identifying when a human needs to step in and notifying the team while keeping the conversation alive.

The value

This automation is the AI receptionist, qualifier, and sales assistant running 24/7 across every channel the business operates on. Every inbound message gets an intelligent, contextually appropriate response within seconds — regardless of the time of day, the volume of enquiries, or who's available on the team. The multi-channel routing means no message falls through regardless of where it comes from. The escalation logic ensures complex or high-value conversations get handed to a human at exactly the right moment, with full context already captured.

Where it fits

This is the frontline of the entire AI communication system — the automation that makes the business feel responsive, professional, and always available without adding headcount. For clients who are missing leads due to slow response times or inconsistent follow-up across channels, this single automation addresses the problem entirely. In sales conversations, this is the most tangible demonstration of what AI-powered business operations actually looks like in practice — not a chatbot, but a fully managed, intelligent conversation system operating across the entire business in real time.

Overall purpose

Automatically manage AI-powered customer conversations by capturing incoming messages, identifying the communication channel, preparing the appropriate AI context, generating responses through external robot webhooks, sending replies through the correct channel, and escalating conversations to a human when the bot cannot respond.

The build — trigger → steps → branches

Triggersentry point

Customer Replied via SMS

Customer Replied

Contact has the tag: new leadMessage channel: SMS

Bot Response Tag Added

Contact Tag Added

Tag added: response

Customer Replied via Instagram

Customer Replied

Message channel: Instagram DM

If / Elsesplits into 3

Who Started the Workflow?

Customer Replied11 steps

Handles a new inbound customer message: stores it, detects the channel, and prepares the AI response.

Store Incoming Message
Action
Update Contact Field · Message Body = {{message.body}}
Store the customer's latest incoming message.
Clear Follow-Up Tag
Action
Remove Contact Tag · robot follow up
Clear any previous automated follow-up status.
Detect Reply Channel
If / Else
Set Response Channel per channel
Records the customer's channel and routes to Add Robot Reply Tag. Facebook = FB; Google Business Profile = GMB; Instagram = IG; SMS = SMS; WhatsApp = WhatsApp; Live Chat = Chat Widget; AI Status Off and None (fallback) take no action.
Add Robot Reply Tag
Action
Add Contact Tag · robot reply
Mark that the next response will be generated by the AI robot.
Check Sales Calendar
If / Else
If Calendar ID empty
If empty, set Calendar ID = {{custom_values.calendar_id}} — the default sales calendar used for booking conversations.
Check Greeting Message
If / Else
If Intro Message empty
If empty, set Intro Message = {{contact.intro_message}} — the default AI greeting message.
Has a Sales Call Been Booked?
If / Else
Tag: sales call | scheduled
Branch on whether the contact has already booked a sales call.
Did Not Book · Update OpenAI Prompt
Action
OpenAI Prompt · "Begin by asking for their name, email, and phone number {{custom_values.openai_prompt}}"
Instruct the AI to collect the contact's details and guide the conversation toward booking a sales call.
Did Not Book · Send Booking Robot Webhook
Action
POST · {{custom_values.robot_webhook}}
Generate a booking-focused AI response.
Booked · Update OpenAI Prompt
Action
OpenAI Prompt · "{{custom_values.openai_prompt}}"
Use the standard AI conversation prompt and avoid requesting contact details that have already been collected.
Booked · Send Non-Booking Robot Webhook
Action
POST · {{custom_values.robot_webhook__nonbooking}}
Generate a general AI response for a contact who has already booked.

Bot Response7 steps

Sends the AI-generated reply on the correct channel, or escalates to a human if the bot cannot answer.

Can the Bot Answer?
If / Else
Answering entity = Bot AND no "alert user" tag
Branches: Answering / Not Answering.
Answering · Clear Response Tag
Action
Remove Contact Tag · response
Clear the tag that triggered the outbound bot-response flow.
Answering · Wait
Wait
Time-Based Wait · 0.01 Minutes
Allow the AI response and contact fields to finish updating.
Answering · Check Response Channel
If / Else
SMS / Instagram / None
Route the outbound reply to the channel the customer used.
Answering · Send SMS
SMS
Send SMS · "{{contact.response}}"
Send the AI-generated response to the customer via SMS.
Answering · Send Instagram DM
Action
Send Instagram DM · "{{contact.response}}"
Send the AI-generated response to the customer via Instagram.
Not Answering · Flag for Human
Action
Remove Tags: alert user, response · Add Tag: alert user
Stop the automated AI response process and flag the conversation for human takeover.

None (Fallback)0 steps

No actions currently configured.

No actions

ai-chatomnichannelwebhookescalationopenai