Hi, I'm
I'm a CS student who learns by building and improving things step by step. I'm especially interested in aviation, backend systems, cloud infrastructure, and machine learning, and I use projects to turn that curiosity into real engineering experience.
I'm a CS student who likes understanding how systems work from end to end. I started with small Python programs, then gradually moved into APIs, databases, cloud deployment, and machine learning by actually building things. My aviation project is the strongest example of that approach, but the main story here is how I learn: pick something real, go deep, and turn it into working software.
I learn best by building things. Here's what I'm actively studying, exploring, and planning to pick up next.
This is the main project that represents how I like to learn: pick a real problem, build the full system, and ship it end-to-end.
A self-learning project built end-to-end — live AirLabs API feed, SQLite snapshot engine, CheckWX METAR weather, Mapbox GL JS satellite map with live flight overlays, and a hybrid XGBoost + rule-engine delay predictor. Built this to learn real-world Python: threading, API design, ML pipelines, and cloud deployment.
Real-time positions for all active Indian flights on a Mapbox satellite map. Auto-refreshes every 10 seconds via a zero-reload JS bridge.
Hybrid XGBoost + rule engine scores every flight for delay risk. Factors: congestion, METAR severity, schedule pressure, route OTP, aircraft type.
Congestion score, inbound density, low-altitude pressure, and 1-hour schedule demand per airport — bottleneck detection before cascades start.
Auto-detects holding patterns, go-arounds, diversions, and stale contacts. Surfaces anomalies before ATC broadcasts them.
Search any active flight by call sign — see live position, estimated arrival, distance to destination, and predicted delay reason.
CheckWX METAR ingestion with deduplication and severity scoring. Layers directly into the delay prediction model and map overlays.
The next aviation-focused idea I want to turn into a real project.