Projects

Built SLO-aware benchmarks and polite LLM throttling on Apple Silicon.

Built a constraint-aware retrieval-augmented QA pipeline for Natural Questions.

Built an on-device, voice-driven YOLO detector for iOS.

Built a real-time Human Activity Recognition (HAR) app for iOS.

Built a noise-robust SVM using KNN and probabilistic weighting.

Built a GUI agent VLM with LoRA-based fine-tuning for user-specific adaptation.

Developed a multi-drone system with autonomous coordination and path planning.

Implemented operating system components in Stanford's Pintos.

Implemented TSP algorithms and proposed an MCMF-based heuristic.

Implemented, benchmarked, and analyzed multiple sorting algorithms.

Explored efficient LLM architectures for on-device environments.

Solved algorithmic problems every day for 366 days, completing 1,000+ problems in 2024.