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

# Platform Overview

> Understand how Chatbot Platform components work together

Chatbot Platform provides a complete infrastructure for building, deploying, and managing AI-powered chatbots across multiple messaging platforms.

## Architecture

The platform consists of four main components that work together:

<CardGroup cols={2}>
  <Card title="Bots" icon="robot" href="/bots/index">
    Your AI chatbot instances with customizable behavior
  </Card>

  <Card title="Integrations" icon="webhook" href="/bots/integrations/index">
    Connections to AI backends (OpenAI, Anthropic, custom APIs)
  </Card>

  <Card title="Channels" icon="plug" href="/bots/channels/index">
    Platform connections (Telegram, Slack, Discord, WhatsApp)
  </Card>

  <Card title="Agent Loops" icon="rotate" href="/agent-loops/index">
    Autonomous AI workflows with tool access via MCP
  </Card>
</CardGroup>

## How Messages Flow

```mermaid theme={null}
graph LR
    A[User on Telegram/Slack] --> B[Channel]
    B --> C[Bot]
    C --> D[Integration Selection]
    D --> E[AI Backend]
    E --> D
    D --> C
    C --> F[Behavior Processing]
    F --> B
    B --> G[User Receives Reply]
```

### Message Flow Steps

<Steps>
  <Step title="User Sends Message">
    A user sends a message through their preferred platform (Telegram, Slack, Discord, etc.)
  </Step>

  <Step title="Channel Receives Message">
    The connected channel receives the message via webhook and validates the request
  </Step>

  <Step title="Bot Processes Message">
    The bot receives the message, retrieves conversation history based on message context settings
  </Step>

  <Step title="Integration Selection">
    If multiple integrations exist, the bot selects one based on A/B testing weights. Otherwise, it uses the default integration.
  </Step>

  <Step title="AI Backend Generates Response">
    The integration forwards the message to your AI backend (OpenAI, Anthropic, or custom webhook). The AI generates a response based on the conversation context.
  </Step>

  <Step title="Behavior Processing">
    The bot applies behavior settings like typing indicators, message detection, and formatting
  </Step>

  <Step title="User Receives Reply">
    The reply is sent back through the channel to the user
  </Step>
</Steps>

## Multi-Tenancy

Chatbot Platform is built with multi-tenancy at its core:

* **Teams**: Create separate workspaces for different projects or clients
* **Isolated Resources**: Each team has its own bots, channels, integrations, and billing
* **Role-Based Access**: Invite team members with specific permissions
* **API Keys**: Each team has unique API credentials for programmatic access

## Key Features

### A/B Testing

Configure multiple integrations per bot with weighted selection. The platform automatically distributes requests across integrations based on your specified weights, allowing you to:

* Compare different AI models
* Test prompt variations
* Gradually roll out new backends
* Implement fallback strategies

Learn more in [A/B Testing](/bots/integrations/ab-testing).

### Conversation Context

Bots maintain conversation history automatically:

* **Message Detection**: Determines which messages belong to the same conversation
* **Context Window**: Configure how many messages to include in each request
* **Platform-Specific Logic**: Thread awareness in Slack, group context in Telegram

Learn more in [Conversations](/bots/conversations/index).

### Typing Indicators

Native or simulated "thinking" indicators across platforms:

* **Native Support**: Real typing indicators on supported platforms (Telegram, Slack)
* **Simulated Messages**: "Thinking..." messages on platforms without native support
* **Configurable**: Enable/disable per bot

Learn more in [Typing Indicators](/bots/behaviors/typing-indicators).

### Agent Loops

Autonomous AI workflows that can:

* Access tools via MCP (Model Context Protocol) servers
* Execute multi-step tasks without user intervention
* Send results to callback webhooks
* Run on scheduled intervals or manual triggers

Learn more in [Agent Loops](/agent-loops/index).

## Common Use Cases

<CardGroup cols={2}>
  <Card title="Customer Support" icon="headset" href="/use-cases/customer-support">
    Automate responses to common questions across all your support channels
  </Card>

  <Card title="Team Assistant" icon="users" href="/use-cases/team-assistant">
    Build internal tools that help your team access information and automate workflows
  </Card>
</CardGroup>

## Next Steps

<CardGroup cols={3}>
  <Card title="Create Your First Bot" icon="rocket" href="/get-started/quickstart">
    Follow our quickstart guide
  </Card>

  <Card title="Explore Bots" icon="robot" href="/bots/index">
    Learn about bot configuration
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/introduction">
    Integrate programmatically
  </Card>
</CardGroup>