Inspiration
As an electronics engineering student fascinated by data systems and real-world applications, I wanted to build something practical that combines my growing programming skills with a universal problem: flight delays. Every time I book flights in India (AI, IndiGo, Vistara), I wonder - how delayed is this flight usually? No website gives this simple insight. This project was born to solve that exact pain point using modern APIs and AI.
What it does
A full-stack web app that: Takes any flight number (e.g., AI202, 6E2174) Fetches live historical data from Aviation Stack API (last 10 flights) Calculates actual delays (scheduled vs actual departure) Uses Google Gemini AI to analyze patterns and deliver traveler-friendly insights:
How we built it
MVP with MongoDB: Started with local database + sample Indian flight data API Pivot: Switched to live Aviation Stack for real global data (no DB needed) AI Integration: Fed flight records to Gemini for natural language analysis Polish: Separated CSS, added responsive design, smooth loading states
Challenges we ran into
API Complexity - Time parsing failed on malformed data → Added robust try/catch + validation Node.js Gotchas - CORS/network errors → Proper middleware + error boundaries Data Quality - 70% of flights missing actual time → Filter + fallback to estimated
Accomplishments that we're proud of
Live data, no database AI-powered insights (not just stats) Separated HTML/CSS for easy customization
What we learned
This project leveled up my full-stack + API + AI skills while solving a real problem I face booking flights. From MongoDB struggles to live Aviation Stack mastery - pure coding growth
Log in or sign up for Devpost to join the conversation.