Projects

End-to-end incident intelligence pipeline using LLMs to ingest logs, generate structured summaries, embed and cluster incidents to identify duplicates and recurring failure patterns. Developed in collaboration with Microsoft, advised by Alison Yao (Microsoft Research).

ML classifiers for ECG arrhythmia detection achieving over 90% accuracy.

Autonomous line-following robot with obstacle detection and touchless activation.

Real-time Whack-a-Mole reaction game on Nexys4 DDR FPGA with multi-difficulty rounds and LED/switch I/O.

Android stock tracking app with real-time data, portfolio persistence, and price alerts.

Arduino-powered temperature monitoring device with real-time alerts and LCD display.