Hi, I'm Tev. I am a Founding Engineer in NYC that builds fun, open-source applied AI projects on the side - small things that are genuinely useful in a workday and a little bit of a pleasure to have installed.

Founding engineer at an early-stage startup · applied AI by day, playful open source by night.

I started out in deep learning research, training CNNs for medical image classification, and drifted steadily closer to the machine: ML systems, cloud infrastructure, and the plumbing that turns a model into something a person can actually use. These days I spend most of my time driving coding agents rather than writing every line myself, which turns out to be its own engineering problem.

The projects here come out of that. I like them close to the machine - local-first, installable in one command, legible without a dashboard - and I like them to have a bit of character. AI is at its best when it disappears into something small and useful, so that is mostly what I ship. More on the way.

• 🐣 parallel-harness-pets - A creature for every git worktree, living in your coding agent's status line. Species comes from the branch name, mood tracks how tidy the branch is. Claude Code, Codex, tmux, any harness. Go

• Bouldering, mostly the falling-off part
• Chasing down new restaurants, then over-recommending them to everyone
• Guitar, electric and acoustic, neither of them quietly
• Spotify playlists I maintain with more rigour than some production systems

• Building small applied AI projects, starting with tooling for parallel agent work
• Learning ML systems, distributed data systems, and high-performance cloud infrastructure
• Curious about GPU programming and how far the local-first version of all this can go

Most of my work lives where applied AI meets software you actually enjoy having around. If you are building in that space - agent tooling, desktop oddities, small models doing useful work - I probably want to see it.