WhatsApp Fact Checker Bot

🎯 Overview

AI-powered WhatsApp bot that detects and debunks misinformation in real-time using Claude and Perplexity APIs. Built to combat the spread of false information in local communities.

🔍 Problem Statement

Misinformation spreads rapidly through WhatsApp groups, especially in local communities where fact-checking resources are limited. This leads to:

Distrust among community members
Poor decision-making based on false information
Social tensions and conflicts

💡 Solution

Built an automated fact-checking bot that:

Receives messages via WhatsApp Business API
Analyzes content using Claude for context understanding
Fact-checks claims using Perplexity’s real-time search
Returns verified information with sources
Handles high-volume traffic through MAKE.com orchestration

🏗️ Architecture

WhatsApp Message → Webhook → MAKE.com → Claude API (Analysis) 
                                       ↓
                              Perplexity API (Fact-check)
                                       ↓
                              Response Processing → WhatsApp Reply

🛠️ Tech Stack

Backend: Python 3.9+
APIs: WhatsApp Business Cloud API, Claude API, Perplexity API
Automation: MAKE.com (workflow orchestration)
Infrastructure: Webhooks, REST APIs
Version Control: Git with unit testing

✨ Key Features

⚡ Real-time fact-checking - Instant responses to user queries
🔄 Queue management - Handles high-volume traffic efficiently
🎯 Context-aware - Claude understands nuanced claims
📊 Source verification - Provides credible sources via Perplexity
🧪 Tested & reliable - Comprehensive unit test coverage

🎬 How It Works

User sends message to WhatsApp bot with a claim or question
Webhook triggers MAKE.com workflow
Claude analyzes the message for key claims and context
Perplexity searches for factual information and sources
Response generated with fact-check results and sources
User receives verified information within seconds

🚀 Impact

Improved information quality in local communities
Reduced spread of misinformation
Increased trust among community members
Scalable solution for fact-checking at scale

🧠 Challenges & Solutions

Challenge 1: Rate Limiting

Problem: WhatsApp API has strict rate limits
Solution: Implemented queue system with MAKE.com to batch process messages

Challenge 2: Context Understanding

Problem: Short messages lack context
Solution: Used Claude’s advanced reasoning to infer context from minimal information

Challenge 3: Source Reliability

Problem: Not all sources are trustworthy
Solution: Perplexity API prioritizes authoritative sources in its search results

📈 Future Enhancements

Multi-language support (Hindi, Marathi, etc.)
Image fact-checking using vision models
User feedback loop for accuracy improvement
Analytics dashboard for tracking misinformation trends
Integration with local news verification services

🔗 Links

GitHub Repository (Coming Soon)
Technical Documentation (Coming Soon)

Built with ❤️ for safer online communities