About

Engineer, researcher, reader, systems-minded generalist

I like difficult technical work that still has to be useful.

My background crosses data science, deep learning, infrastructure, and software engineering. I do my best work on problems that need both technical depth and practical implementation.

Working style

How I tend to approach problems

  • I prefer explicit systems over magical ones. Good abstractions are earned by surviving contact with edge cases.
  • I care about speed, but the real gains usually come from better interfaces, fewer handoffs, and cleaner feedback loops.
  • I like reading source code, tracing data movement, and understanding failure modes before trying to "improve" anything.
  • I am unusually comfortable moving between algorithmic work and operational plumbing.

Interests

Things I keep coming back to

Technical

Efficient ML and kernel-level thinking

Long-context methods, hardware-aware optimization, and performance work that translates into tangible throughput wins.

Systems

Pipelines that become easier to trust

Better orchestration, test coverage, and tooling so teams spend less time fighting the system around the model.

Writing

Essays that use technical ideas as explanatory tools

I like using concepts from optimization, economics, and complex systems to think clearly about social and technological change.