Documentation Index
Fetch the complete documentation index at: https://docs.chatzy.ai/llms.txt
Use this file to discover all available pages before exploring further.
What Are Squads?
A Squad is a team of multiple Conversational AI Agents that work together to handle complex workflows and multi-step processes.Instead of relying on a single generalist agent, a Squad divides tasks among specialized agents each with its own role, expertise, and knowledge base, ensuring faster, more accurate, and scalable automation.
Best For
- Multi-stage customer workflows (lead → booking → support)
- Departmental automation (sales, support, operations)
- Escalation and handoff between specialized agents
- AI-driven collaboration for context-rich conversations
Key Strengths
- Collaborative Intelligence: Multiple agents coordinate to complete end-to-end workflows.
- Role-Based Specialization: Each agent focuses on a defined domain - e.g., Lead Qualification, Booking, Support.
- Shared conversation history: When the conversation is getting transfered between persitent agents, they share the conversation history.
If your workflow is large and involves multiple API calls, there’s a good chance users may ask questions that fall outside the responsibility of the current agent. In such cases, Squad is an ideal solution. For example, imagine you run a car-rental business with an AI Sales agent that handles pricing, introductions, and general sales queries, and a separate Service agent that manages billing and onboarding. If, during onboarding, the user suddenly asks to change the car, the conversation needs to be handed back to the Sales agent. Squad automatically manages this kind of agent handoff, ensuring the user is always routed to the right agent without manual intervention.
Squad Architectures
You can choose from the following multi-agent architectures, depending on your use case and workflow design Network
Supervisor
Custom
| Architecture | Description |
|---|---|
| Network | All agents communicate freely with each other, ideal for collaborative brainstorming or multi-domain reasoning. |
| Supervisor | A central controller agent assigns tasks to others and consolidates results, ensuring structured flow. |
| Custom | Build your own connection graph between agents to match unique process logic and data flow. |
Create a Squad
Build your first Squad and enable collaborative AI workflows.
Before creating the Squad, you must properly configure all the required Conversational AI agents and clearly specify in their prompts the conditions under which a conversation should be transferred from one agent to another. For example:
Once you create a Squad, you’ll see the following screen:
