I'm an AI safety researcher and software engineer with a focus on technical alignment research. My background spans data platforms, ML systems, 3D visualization, and computational geometry.
A novel research agenda to impose just enough structure on latent embeddings to make AI more interpretable and steerable.
Paper @ ICLR 2026 • notebooks • source • articles
Novel techniques for detecting out-of-distribution inputs in language models: a new visualization and metric.
A library and template repository to quickly spin up new AI research projects, with sensible defaults and access to cloud GPU compute.
A slick visual timer app.
A unique game featuring innovative physics for character movement and environment interaction.
Contributions to Blender
- Physically-based "classical" SPH solver
- Adaptive time-step for fluid particles
- Baking (rendering) to vertex colors
- AI safety: technical alignment research
- ML engineering and deployment pipelines
- Data platforms with robust governance
- 3D graphics, game development, simulation, computational geometry, visualization
- Web development (full-stack, various frameworks)
- Information retrieval and search systems
- Systems design




