Seamlessly bridge WhatsApp → LLM → Notion to automate tracking of internship and placement alerts. Built using Node.js, whatsapp-web.js, Ollama (Mistral), and Notion SDK.

This project was built to automate the organization and tracking of internship and placement opportunities shared by my college’s Training & Placement cell on WhatsApp. Managing these messages manually was inefficient and error-prone, so this backend streamlines the process by parsing messages with LLMs and syncing structured data to Notion.

• ✅ Real-time WhatsApp message ingestion via whatsapp-web.js
• 🧠 Local LLM parsing using Ollama with models like mistral
• 🗂️ Structured data sync to Notion using Notion SDK
• 🔁 Message queueing and retry logic for stability
• 🔐 Environment-based configuration
• 🧪 Easily testable with POST endpoint (/parse)
• 🔍 Flexible message listening with detailed message logging

1. Check logs/whatsapp_message_log.json after receiving messages to find the WhatsApp IDs of your desired contacts or groups.
2. Edit services/whatsappService.js and add those IDs to the targetContact array:

   const targetContact = [
     '[email protected]', // Example contact
     '[email protected]', // Example group
     // Add more as needed
   ];
   

• Backend: Node.js, Express.js
• WhatsApp Integration: whatsapp-web.js
• LLM Parsing: Local LLM (e.g., Mistral) via Ollama
• Database Sync: Notion SDK
• Others: dotenv, axios, cors, nodemon

• Node.js
• Ollama installed and running locally
• Notion integration + database (follow Notion SDK setup)
• WhatsApp account [ The qr will be displayed in the terminal when you run the project ]

git clone https://github.com/avinash7r/WhatsApp-Notion-sync.git
cd interntrack-backend

npm install

Create a .env file based on .env.example:

Then fill in:

• NOTION_API_KEY – Get your integration key
• NOTION_DB_ID – Your Notion database ID
• OLLAMA_URL – http://localhost:11434/api/chat
• Any other relevant vars

Ensure Ollama is installed and running:

ollama pull mistral
ollama run mistral

npm run dev

This project is licensed under the MIT License.