Technical notes from my AI, ML, and full-stack engineering journey.
How I built an end-to-end deep learning pipeline that detects PCB manufacturing defects with 98.3% recall, processing boards at industrial speeds.
How I built AigilityX to combine intelligent agents, voice interaction, RAG, and live WebSocket communication into one production-style platform.
Implementing a deep learning system for camouflaged object detection using PyTorch and SINet, tackling one of computer vision's hardest problems.
How I built a fully automated, serverless MLOps pipeline that forecasts AQI three days ahead with explainability, alerts, and a live dashboard.