> ## Documentation Index
> Fetch the complete documentation index at: https://docs.conversimple.com/llms.txt
> Use this file to discover all available pages before exploring further.

# The Agent

> Learn about the ConversimpleAgent session model.

Each `ConversimpleAgent` instance handles a single conversation session. For multiple concurrent conversations, you should create multiple agent instances. This one-to-one mapping ensures that conversation state is managed cleanly and that each conversation has its own isolated process.

## Agent Lifecycle

The lifecycle of an agent is straightforward:

1. **Initialization**: You create an instance of your agent class, which inherits from `ConversimpleAgent`.
2. **Connection**: The `agent.start()` method establishes a WebSocket connection to the Conversimple platform.
3. **Ready**: The agent is ready to handle a conversation.
4. **Termination**: The `agent.stop()` method disconnects the agent from the platform.

Here is an example of how to manage agent instances for multiple conversations:

```python theme={null}
# Per-conversation agent instances
async def handle_conversation(conversation_id):
    agent = MyAgent(api_key=api_key, customer_id=customer_id)
    await agent.start(conversation_id=conversation_id)
```

This model of one agent per conversation is recommended for most use cases and simplifies state management.