
An OpenPyXL-compatible spreadsheet engine with native Rust readers, calculation, and rendering, exposed through a Python API.
Open to new roles · San Francisco or remote

An OpenPyXL-compatible spreadsheet engine with native Rust readers, calculation, and rendering, exposed through a Python API.
Live AI model landscape dashboard tracking model quality, pricing, speed, benchmark trust, and leaderboard movement across providers.

Interactive time-lapse map of San Francisco metered parking activity, built from 206M+ meter transactions and shared through a viral X demo.
A benchmark that scores web-search providers (Exa, Perplexity, Tavily, Gemini, Grok) on everyday research quality and — the differentiator — landscape recall: whether a search surfaces every real competitor in a market, including the obscure long-tail that naive agent fan-out misses.
Drive a real, logged-in Chrome profile from local scripts via a Native Messaging bridge. No remote-debugging popup, no separate browser profile.
Want to see more?
44+ merged pull requests across the agent-tooling ecosystem: terminal rendering, performance work, and UX in the tools I use every day.
The agent development environment for running fleets of parallel coding agents. Contributions span terminal rendering, performance, and workspace UX.
AI coding agent for the terminal. Contributions across the coding-agent core, TUI, and provider layer: caching, mouse routing, and correctness fixes.
Multi-provider AI proxy. Contributed model registry metadata for new frontier models.
Pure-Rust spreadsheet library for reading and writing XLSX. Contributed upstream fixes discovered while building high-performance Excel tooling.

I started in finance at PwC CMAAS, advising on IPOs and complex accounting and automating data pipelines for aerospace CFOs. That work taught me what high-stakes workflows demand: clear ownership, traceable evidence, and software people can trust.
Now I build developer tools, applied AI systems, and data products in Rust, Python, and TypeScript. Recent work spans a spreadsheet engine with native calculation and rendering, evidence-gated agent workflows, open-source tooling, and production data visualizations.
I work end to end, from architecture and verification through release engineering and the interface a user touches. I am looking for a team where systems depth, product ownership, and experience with high-stakes workflows are an edge.
Building an applied AI product, developer tool, or system where correctness matters? Let’s talk.