Skip to main content
HoopAI allows you to create and manage multiple Conversation AI bots within a single account. Instead of relying on one bot to handle every scenario, you can deploy specialised bots for different departments, languages, campaigns, or customer segments — then orchestrate handoffs between them using workflows. This guide covers the architecture behind multi-bot setups, when to use them, and how to configure seamless bot-to-bot transfers.
Before diving into multi-bot orchestration, make sure you are comfortable with Conversation AI basics and bot settings, including the distinction between primary and non-primary bots.

​
Primary vs non-primary bots

Understanding the difference between primary and non-primary bots is the foundation of multi-bot orchestration.

​
Primary bot

  • There can be only one primary bot per account at any given time.
  • The primary bot handles all inbound conversations on its assigned channels unless a workflow has explicitly assigned a different bot to that contact.
  • Think of the primary bot as your “default responder” — it catches everything that is not routed elsewhere.
  • The primary bot is configured in bot settings by toggling the Primary flag.

​
Non-primary bots

  • Non-primary bots are specialist bots that only activate when triggered by a workflow.
  • They do not respond to inbound messages on their own — they must be explicitly assigned to a contact through a workflow action.
  • You can have unlimited non-primary bots, each with its own prompt, knowledge base, and configuration.
  • Non-primary bots are ideal for department-specific tasks, language routing, or campaign-specific conversations.
FeaturePrimary botNon-primary bot
Responds to inbound messages automaticallyYes (on assigned channels)No (must be triggered by workflow)
Number allowed per accountOneUnlimited
Channel assignmentRequiredOptional
Activated byIncoming messageWorkflow action
Best forGeneral inquiries, default handlingSpecialised tasks, campaigns, departments

​
When to use multiple bots

A single bot works well for many businesses. But as your operations grow in complexity, multi-bot orchestration becomes essential. Here are the most common scenarios.

​
Different departments

Create separate bots for sales, support, and billing. Each bot has its own prompt tailored to the department’s goals, tone, and knowledge base.
  • Sales bot: Focused on qualifying leads, answering product questions, and booking demo appointments.
  • Support bot: Trained on troubleshooting guides and FAQ content, focused on resolving issues.
  • Billing bot: Handles payment inquiries, invoice questions, and subscription changes.

​
Multiple languages

If your customer base spans multiple languages, create a bot for each language. Each bot’s prompt and knowledge base should be written entirely in the target language.
  • English bot: Primary bot handling the default language.
  • Spanish bot: Non-primary bot triggered when the contact’s language preference is Spanish.
  • French bot: Non-primary bot for French-speaking contacts.
Use a workflow to detect the contact’s language (via a custom field or the language of their first message) and route them to the appropriate bot automatically.

​
Campaign-specific bots

When running targeted marketing campaigns, create a dedicated bot for each campaign. This allows you to:
  • Tailor the conversation to the campaign’s offer and messaging.
  • Use a campaign-specific knowledge base with relevant landing page content.
  • Track bot performance per campaign on the Conversation AI dashboard.

​
Funnel stages

Use different bots at different stages of your sales funnel:
  1. Top of funnel: A friendly, inquisitive bot that qualifies leads and collects basic information.
  2. Middle of funnel: A knowledgeable bot that provides detailed product information and handles objections.
  3. Bottom of funnel: A closing bot focused on booking appointments, sending proposals, or processing orders.

​
Agency multi-client management

If you are an agency managing multiple clients within sub-accounts, you can create client-specific bots within each sub-account. For agencies managing at the parent level, you can configure bot templates that are deployed across client accounts with customised prompts and knowledge bases for each client.

​
Bot transfer workflows

Bot transfers are the mechanism that makes multi-bot orchestration possible. A transfer moves a conversation from one bot to another using workflow actions.

​
How bot transfers work

When a transfer occurs, the following happens:
  1. The current bot stops responding to the contact.
  2. The new bot is assigned to the contact and begins handling the conversation.
  3. The conversation history is preserved — the new bot can see previous messages for context.
  4. The contact’s bot status is updated to reflect the new bot assignment.

​
Setting up a bot transfer workflow

1

Create or open a workflow

Navigate to Automation > Workflows and create a new workflow or open an existing one.
2

Choose a trigger

Select the appropriate trigger for when the transfer should occur. Common triggers include:
  • Contact tag added (e.g., “needs-support” tag triggers transfer to support bot)
  • Conversation AI action fired (the current bot triggers a custom action)
  • Custom field updated (e.g., language preference changes)
  • Form submitted (a campaign form submission triggers a campaign bot)
3

Add the 'Assign to Bot' action

In the workflow actions, add an Assign to Conversation AI Bot action. Select the target bot from the dropdown.
4

Set the bot mode

Choose whether the new bot should operate in Auto-Pilot or Suggestive mode for this contact.
5

Optional: add a handoff message

Add a Send Message action before the bot assignment to inform the customer about the transfer. For example: “Let me connect you with our support specialist who can help with that.”
6

Test the workflow

Trigger the workflow with a test contact and verify the new bot takes over the conversation correctly.

​
Transfer triggered by bot actions

One of the most powerful patterns is having the bot itself trigger its own transfer. Using Conversation AI actions, you can configure the bot to fire a custom action when it detects that the conversation needs a different specialist. For example, your primary sales bot could be prompted with: 
If the customer asks about billing, refunds, or payment issues,
trigger the "transfer-to-billing" action. Do not attempt to answer
billing questions yourself.
The “transfer-to-billing” action then triggers a workflow that assigns the billing bot to the contact.

​
Architecture patterns

Below are common multi-bot architectures. Choose the one that best fits your business.

​
Pattern 1: Hub and spoke

The primary bot acts as a “receptionist” that triages incoming conversations and routes them to specialist bots. 
Incoming message
      |
      v
 Primary bot (triage)
      |
   --------+--------+---------
   |                |         |
   v                v         v
Sales bot    Support bot   Billing bot
How it works: The primary bot asks the customer what they need help with, then triggers the appropriate action to transfer to the specialist bot. Each specialist bot has its own prompt and knowledge base. Best for: Businesses with clearly defined departments and high conversation volume.

​
Pattern 2: Language router

A workflow detects the contact’s language and assigns the appropriate language-specific bot before any conversation begins. 
Incoming message
      |
      v
 Language detection workflow
      |
   --------+--------+---------
   |                |         |
   v                v         v
English bot   Spanish bot   French bot
How it works: A workflow trigger fires on the first inbound message. A condition checks the contact’s language custom field (or uses keyword detection on the message). The contact is then assigned to the corresponding language bot. Best for: Businesses serving international markets or multilingual communities.

​
Pattern 3: Campaign cascade

Different marketing campaigns route to different bots, each tailored to the campaign’s offer and audience. 
Campaign A form  -->  Campaign A bot
Campaign B SMS   -->  Campaign B bot
Campaign C ad    -->  Campaign C bot
         |                |
         v                v
    (all eventually hand off to)
              |
              v
        Primary bot (general)
How it works: When a contact enters through a specific campaign (via form submission, ad click, or trigger link), a workflow assigns them to the campaign-specific bot. Once the campaign conversation concludes, the bot triggers an action that reassigns the contact to the primary bot for ongoing communication. Best for: Marketing-heavy businesses running multiple concurrent campaigns.

​
Pattern 4: Funnel progression

Bots are assigned sequentially as the contact moves through pipeline stages. 
New lead  -->  Qualifier bot
                    |
             (lead qualified)
                    |
                    v
             Nurture bot
                    |
              (ready to buy)
                    |
                    v
              Closer bot
How it works: Pipeline stage changes (tracked via opportunities or custom fields) trigger workflows that reassign the contact to the next bot in the sequence. Each bot’s prompt and goals align with that stage of the journey. Best for: Businesses with well-defined sales processes and longer sales cycles.

​
Managing bot status per contact

Each contact has a bot status that determines how the Conversation AI interacts with them. Understanding and managing this status is critical for multi-bot setups.

​
Bot status values

  • Active: The assigned bot will respond to messages from this contact.
  • Paused: The bot is temporarily paused. It will not respond but can be resumed.
  • Stopped: The bot has been stopped for this contact (either manually or via a stop bot action). It will not respond until re-activated.

​
Managing status in workflows

You can control bot status through workflow actions:
  • Enable bot for contact: Sets the bot to Active for the contact.
  • Disable bot for contact: Stops the bot for the contact.
  • Assign bot: Assigns a specific bot and sets it to Active.

​
Status in the conversations inbox

You can view and manually change a contact’s bot status directly in the conversations inbox. Look for the bot status indicator in the contact details panel. You can also filter conversations by bot status — see bot status management for details.
When transferring between bots, always ensure the previous bot is stopped before the new bot is activated. Otherwise, both bots may attempt to respond to the same message. The Assign to Bot workflow action handles this automatically, but if you are building custom logic, account for this explicitly.

​
Multi-bot strategy for agencies

Agencies managing multiple client accounts benefit greatly from multi-bot orchestration. Here are strategies to scale effectively.

​
Template-based bot creation

Create bot prompt templates for common industries (dental, real estate, home services, etc.) and deploy them across client sub-accounts. Customise the knowledge base and business details for each client while keeping the core prompt structure consistent.

​
Centralised monitoring

Use the Conversation AI dashboard within each sub-account to monitor bot performance. Track key metrics like:
  • Response rate per bot
  • Appointment booking rate
  • Escalation rate
  • Customer satisfaction indicators

​
Standard operating procedures

Document your multi-bot architecture for each client. Include:
  • Which bots are deployed and their purposes
  • What triggers bot transfers
  • Escalation paths to human agents
  • Knowledge base update schedules

​
Scaling considerations

  • Naming conventions: Use clear, consistent bot names across accounts (e.g., “Sales Bot - [Client Name]”, “Support Bot - [Client Name]”).
  • Knowledge base hygiene: Regularly review and update each bot’s knowledge base to keep information accurate.
  • Prompt versioning: Keep track of prompt changes so you can roll back if a new prompt performs worse.

​
Best practices for multi-bot orchestration

  1. Start simple. Begin with a primary bot and add specialist bots only when you have a clear need. Over-engineering a multi-bot setup adds complexity without proportional benefit.
  2. Define clear boundaries. Each bot should have a well-defined scope. Overlapping responsibilities lead to confused routing and inconsistent customer experiences.
  3. Use handoff messages. When transferring between bots, send a brief message to the customer explaining the transfer. This prevents confusion when the tone or style changes.
  4. Monitor transfer rates. If a bot is transferring too many conversations, its prompt or knowledge base may need refinement. Use the dashboard to track this.
  5. Test end to end. Before deploying a multi-bot setup, test every transfer path with real conversation scenarios. Verify that contacts move smoothly between bots.
  6. Keep knowledge bases separate. Each bot should have its own knowledge base focused on its domain. Shared knowledge bases can cause irrelevant information to surface.

​
Next steps

Conversation AI overview

Review the fundamentals of setting up and configuring a Conversation AI bot.

Conversation AI actions

Configure bot actions including transfers, escalation, and appointment booking.

Bot settings

Manage primary and non-primary bot assignments and channel configuration.

Multi-channel deployment

Deploy your bots across SMS, Instagram, Facebook, WhatsApp, and web chat.
Last modified on March 5, 2026