Systems Software

Privacy-First Homework Help App

Build and test an offline homework-help app with translation and math OCR, then measure accuracy, speed, and privacy tradeoffs.

LLM Agent Reliability Benchmark

Build a benchmark for agentic tool use in CLI tasks, then measure reliability, repeatability, and scoring accuracy with reproducible scripts.

Smartphone Sign Language Tutor

Build and test a smartphone sign-language tutor, measure gesture scoring, and compare adaptive feedback against a static-video baseline.

Shopping Cart