OBJECTIFE - AI-Powered News Synthesis Platform

Inspiration

In an era of information overload and media bias, I was inspired to create a platform that could synthesize real-time news with complete objectivity. The goal was to leverage AI to provide factual, unbiased reporting while maintaining journalistic integrity.

What I Learned

  • AI Integration: Mastered integrating Perplexity AI for real-time news synthesis
  • Full-Stack Development: Built a complete TypeScript/React application with Supabase backend
  • Edge Functions: Deployed serverless functions for news processing and API integrations
  • JSON Parsing: Implemented robust parsing strategies for handling AI-generated content
  • Image APIs: Integrated Pexels API for dynamic, content-relevant imagery

How I Built It

  • Frontend: React + TypeScript with Tailwind CSS for responsive design
  • Backend: Supabase with PostgreSQL database and Edge Functions
  • AI Integration: Perplexity API for premium news synthesis
  • Image Processing: Pexels API with smart keyword extraction
  • Architecture: Serverless functions for scalable news generation

Key Features

  • Real-time news synthesis from multiple sources
  • Objectivity scoring for unbiased reporting
  • Dynamic image matching based on article content
  • Category-specific news filtering
  • Location-based news targeting

Challenges Faced

  1. JSON Parsing Complexity: AI responses required multiple fallback parsing strategies
  2. Image Relevance: Ensuring each article gets unique, content-specific images
  3. API Rate Limits: Managing multiple API calls efficiently
  4. Real-time Processing: Balancing speed with content quality
  5. Error Handling: Building robust fallbacks for API failures

Technical Stack

  • React, TypeScript, Tailwind CSS
  • Supabase (PostgreSQL, Edge Functions)
  • Perplexity AI, Pexels API
  • Deno runtime for serverless functions

The result is a platform that delivers objective, AI-synthesized news with relevant imagery, providing users with factual information free from editorial bias.

Built With

Share this project:

Updates