Lesson 1 of 4
One Bot, All Platforms
Estimated time: 3 minutes
One Bot, All Platforms
In this course, you'll deploy a single AI assistant that works across WhatsApp, Telegram, Slack, and Discord — all at the same time, with consistent behavior and personality.
Prerequisite
Make sure you've completed Getting Started with OpenClaw before this course. You need OpenClaw installed, the Gateway running, and at least two chat platforms connected.
The Problem
Most businesses and teams juggle multiple chat platforms. Your customers are on WhatsApp, your team is on Slack, and your community is on Discord. Running a separate bot for each means:
- Duplicated effort — maintaining the same logic in multiple places
- Inconsistent responses — each bot drifts over time
- More things to break — every platform is a separate point of failure
The Solution
OpenClaw's multi-channel architecture lets you define one assistant — one personality, one set of skills, one knowledge base — and connect it to every platform simultaneously. A message on WhatsApp gets the same quality response as one on Slack.
Architecture Overview
┌──────────────┐
┌────>│ WhatsApp │
│ └──────────────┘
┌──────────────┐ │ ┌──────────────┐
│ OpenClaw │──┼────>│ Telegram │
│ Gateway │ │ └──────────────┘
│ │ │ ┌──────────────┐
│ One Agent │──┼────>│ Slack │
│ One Config │ │ └──────────────┘
└──────────────┘ │ ┌──────────────┐
└────>│ Discord │
└──────────────┘
The Gateway handles all the platform-specific details — message formats, media types, rate limits — so your assistant logic stays clean and portable.
Each platform has an adapter inside the Gateway. When a message arrives on WhatsApp, the WhatsApp adapter converts it into a standard format, passes it to your agent, and converts the response back into WhatsApp's format.
This means your agent never deals with platform-specific APIs. It just receives text (and optional media) and returns a response.
Some features are platform-specific — Slack threads, Discord reactions, WhatsApp quick-reply buttons. OpenClaw provides optional hints your agent can use:
platform— which platform the message came fromthread_id— for threaded conversations (Slack, Discord)capabilities— what the platform supports (buttons, rich media, etc.)
Your agent can use these hints or ignore them. Either way, the basic text response works everywhere.
Real-World Example
A small bakery runs one OpenClaw assistant that handles orders on WhatsApp, answers FAQs on their website chat, and manages team communications in Slack — all with consistent responses and the same friendly tone.
Course Structure
| Lesson | What You'll Do | Time |
|---|---|---|
| 1. One Bot, All Platforms | You are here — understand the architecture | 3 min |
| 2. Connecting Chat Channels | Connect WhatsApp, Telegram, Slack, or Discord | 5 min |
| 3. Setting Assistant Personality | Define tone, knowledge, and behavior rules | 5 min |
| 4. Testing & Deploying | Test across platforms and go live | 5 min |
Lessons 1-2 are free
You can have a working multi-channel bot by the end of Lesson 2. Lessons 3 and 4 teach you how to give it a polished personality and deploy it reliably.
Quick Check
What is the main advantage of OpenClaw's multi-channel approach?
Advanced Message Routing Across Channels
# Routing Like a Pro with OpenClaw Multi-channel routing is the backbone of a sophisticated AI agent infrastructure. It allows you to direct information flow based on origin, sender, and content, ens...