Frontend systems
React, TypeScript, SSR, accessibility, state cleanup, and route-level behavior that survives production.
Bring me the part your team keeps avoiding.
The ugly route. The hydration bug. The Rust service nobody wants to profile. The AI workflow that works until it matters. I am useful when the problem needs sustained focus and fewer polite guesses.
Send the technical problem
Stack trace, repo notes, and constraints are welcome.
The work
Backend and performance
Rust, Node, data-heavy workflows, latency cuts, deployment seams, and boring reliability.
AI product engineering
Agent runtimes, model routing, tool contracts, RAG, MCP, and evaluation loops that behave like software.
How I slot in
- 01
Read the failing boundary before touching the obvious file.
- 02
Make the smallest useful fix that changes the actual outcome.
- 03
Leave behind sharper code, sharper tests, and sharper ownership.
Good fit
- You need staff-level individual contribution without ceremony.
- You have a technical problem with a real business consequence.
- You want direct communication and defensible tradeoffs.
Bad fit
- You need someone to disappear into process theater.
- You want architecture astronauts with no shipped code.
- You cannot tolerate hearing the obvious part out loud.
I am not cheap. I am cheaper than the wrong fix.
Send the hard part. If I am the wrong person, I will say that quickly.