Full-stack web app democratizing college advice with semantic search for 200+ opportunities. Scaled to 500 users in 5 days with 100K LinkedIn impressions. Built with Grok 3, Xenova embeddings, and Vercel.
Developing a transformer model in Rust from scratch. Created data pipeline and processing for pretraining. Focus on understanding low-level implementation details.
Chrome extension analyzing route safety using TensorFlow-based ML backend. Combines live-time and historical data for safety analysis rather than speed optimization.
Next.js app analyzing news bias and surfacing balanced perspectives. Built user interface and backend to identify source disagreement using NLP methods.
Collaborated with Harvard PhDs to develop memory layers for quantum sensing automation. Exploring ML applications in quantum computing research.
Trained Neural Network from scratch using basic Python packages for data management. Built to understand fundamental ML concepts without high-level frameworks.
Authored regression analysis of inflation vs. soldier pay in Late Antiquity. Featured by the National Committee for Latin and Greek.
