Trained a Discrete Denoising Diffusion Probabilistic Model (D3PM) on 500,000 chess puzzle positions — with no explicit knowledge of the rules. The model achieves 69.5% valid position generation versus a 16.3% random baseline, and closely matches the training distribution on pawn structure metrics.
Built a multi-agent RAG system using IBM watsonx Orchestrate to help UB Computer Science students navigate coursework, degree requirements, and faculty research — delivering personalized responses across a range of academic queries. Placed 2nd at IBM's Agentic AI Hackathon at UB.
Office Hours are important for the academic success of many students, but often end up being hectic and inefficient. This project focuses on streamlining the office hours experience for both students and instructors.
Building...