Skip to main content
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:

Bots

Your AI chatbot instances with customizable behavior

Integrations

Connections to AI backends (OpenAI, Anthropic, custom APIs)

Channels

Platform connections (Telegram, Slack, Discord, WhatsApp)

Agent Loops

Autonomous AI workflows with tool access via MCP

​
How Messages Flow

​
Message Flow Steps

1

User Sends Message

A user sends a message through their preferred platform (Telegram, Slack, Discord, etc.)
2

Channel Receives Message

The connected channel receives the message via webhook and validates the request
3

Bot Processes Message

The bot receives the message, retrieves conversation history based on message context settings
4

Integration Selection

If multiple integrations exist, the bot selects one based on A/B testing weights. Otherwise, it uses the default integration.
5

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.
6

Behavior Processing

The bot applies behavior settings like typing indicators, message detection, and formatting
7

User Receives Reply

The reply is sent back through the channel to the user

​
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.

​
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.

​
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.

​
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.

​
Common Use Cases

Customer Support

Automate responses to common questions across all your support channels

Team Assistant

Build internal tools that help your team access information and automate workflows

​
Next Steps

Create Your First Bot

Follow our quickstart guide

Explore Bots

Learn about bot configuration

API Reference

Integrate programmatically