Yousuf Golding
Machine Learning Engineer · NLP Researcher
I build and evaluate LLM systems across two threads I care about equally: making them safer and measurable, and using them to make learning better.
Safety & Evaluation
LLM evaluation methodology, export-control tooling, and agent-safety monitoring.
Educational AI
Conversational tutoring, voice-first learning, and QUD-grounded co-writing.
Systems & Tooling
Multi-drone search and rescue, and local-first developer tooling.
M.S. in Natural Language Processing (UC Santa Cruz) and B.S. in Applied Mathematics. Recent work spans multi-agent AI-safety monitoring, applied ML tooling, and conversational AI for education. Earlier: NLP research internships and four years as a science educator at Chabot Space & Science Center and NASA Ames.
Featured
All work →
Systems & Tooling SARchlight - Multi-Drone Search and Rescue
Terrain-aware multi-drone search with a Bayesian belief map, a live dashboard, and a deployed voice agent. Built with a team at a search-and-rescue hackathon.
View project →
Safety & Evaluation Dual-Use Navigator
A cross-jurisdictional reference tool that compares dual-use export-control entries across US, EU, and Australian control lists side by side.
View project →
Educational AI · Research QUD-Empowered LLM Essay Co-writing Tool
Full-stack essay co-writing platform grounded in the Questions Under Discussion (QUD) framework, pairing LLM-generated questions with a student-facing study workflow.
View project →