Automating WhatsApp Message Forwarding to a Channel with n8n, Docker, and WAHA

In this post, I share how I automated the forwarding of messages from a WhatsApp group to a channel using n8n, Docker, and the WAHA library. I set everything up on an old laptop I use as a local server. The goal is to make it easier to manage a ridesharing group between Zamora and Salamanca, which had reached its participant limit.

Challenge Overview

The WhatsApp group "Zamora-Salamanca" connects drivers and passengers to share rides between Zamora and Salamanca.
After reaching the 1000-participant limit, a WhatsApp channel was created, but forwarding messages was manual and tedious.
The challenge: automate filtering and forwarding of ride posts, including the phone number, and mark forwarded messages.

Technical Solution

General Architecture

WAHA: exposes a WhatsApp API in Docker to automatically receive messages.
n8n: orchestrates the workflow, filters messages, and forwards them to the channel.
Google Gemini: free AI to distinguish ride posts from other messages.

Automated Workflow

Message Reception
WAHA receives messages from the group and sends them to an n8n webhook.
Smart Filtering
n8n queries Google Gemini to decide if the message is a ride post or not.
It also automatically discards messages with keywords like "someone" or "looking for".
Enrichment
If the phone number is missing, the system adds it automatically.
Forwarding
The message is published in the WhatsApp channel.
Visual Confirmation
The bot adds a robot reaction 🤖 to the original message to indicate it was forwarded successfully.

Issues and Lessons Learned

WAHA Disconnection
The server restarted every Saturday at dawn and WAHA stopped sending messages. I solved it with an automatic Docker container restart routine.
Incorrect Message Filtering
At first, questions or replies were forwarded. I added a keyword filter and improved the AI prompt to reduce false positives.

Screenshots

n8n Workflow

Code and Configuration

Example docker-compose.yml

# filepath: ./docker-compose.yml
version: '3'
services:
  waha:
    image: devlikeapro/whatsapp-http-api
    restart: always
    ports:
      - "3000:3000"
    environment:
      - WHATSAPP_API_KEY=XXXX
      - WAHA_DASHBOARD_USERNAME=XXXX
      - WAHA_DASHBOARD_PASSWORD=XXXX
      - WHATSAPP_DEFAULT_ENGINE=WEBJS
      - WAHA_PRINT_QR=False
      - WAHA_MEDIA_STORAGE=LOCAL
      - WHATSAPP_FILES_LIFETIME=0
      - WHATSAPP_FILES_FOLDER=/app/.media
      - WHATSAPP_RESTART_ALL_SESSIONS=True
      - WAHA_AUTO_START_DELAY_SECONDS=0
      - WHATSAPP_HOOK_URL=https://MY-DOMAIN/webhook/22ad9ac5-be66-XXXX-XXXX-XXXXXXXXX/waha
      - WHATSAPP_HOOK_EVENTS=message,state.change
  n8n:
    image: n8nio/n8n
    restart: always
    ports:
      - "5678:5678"
    environment:
      - N8N_BASIC_AUTH_ACTIVE=true
      # ...otras variables necesarias...
    volumes:
      - ./n8n-data:/home/node/.n8n

Technical Considerations

The server is an old laptop running Linux and Docker.
The workflow is robust against restarts and network errors.
Filtering can be improved by adding new keywords or adjusting the AI prompt.

Future Improvements

Improve filtering using AI and user feedback.
Add logs and alerts in case the bot fails.
Automate updating the keyword list.

Final Thoughts

This automation has reduced manual effort and made the group much more useful for everyone. If you have a large group and need to automate repetitive tasks, I encourage you to try n8n and Docker.

Would you like to see the full n8n workflow or need help configuring WAHA? Leave your questions in the comments.