CS grad student at UofL focused on AI systems, applied ML, and building things that work outside of a notebook. Currently finishing my M.S. while doing federated learning research and shipping LLM applications for clinical use.
Languages
ML & AI
Web & Systems
Mobile
Agentic tutoring platform — not just a chat interface. Real-time voice interaction, recursive sub-agent tool use (code runners, diagram generators), and a full course publishing/enrollment model. Live in production with 166+ passing tests.
LLM-powered clinical documentation tool built with medical professionals at UofL. Handles the paperwork so neonatal care workers can focus on patients.
Research into federated learning aggregation. Proposed a novel method using unsupervised clustering on gradient properties. Ongoing work at UofL.
AI security research evaluating whether perturbation-based tools actually protect artists from style-mimicry models. Tested a 4×4 matrix of protections vs. countermeasures on Stable Diffusion 3.5 Large, with both algorithmic benchmarks and a human evaluation study.
Fine-tuned lightweight CNNs to classify mushroom species above ImageNet benchmarks, then wrapped it in a Flutter app that runs inference fully on-device — no network needed.
Raspberry Pi + object detection to catch garden pests and fire a solenoid valve to scare them off. Chipmunks hate it.
CLI tool that turns lecture videos into HTML study notes using LLMs. Built it for myself, works well enough to share.
University of Louisville — M.S. Computer Science (Expected May 2026)
Speed School of Engineering · Coursework: AI, Advanced AI, AI Security, Data Mining
CompTIA
- Email: richard.douglas@louisville.edu
- LinkedIn: richard-douglas-a99a00271