Fraud and internal audit data analyst by day. Systems builder and open source maintainer by night, with my projects living under Sigilweaver.

My background is originally in the biosciences, but I pivoted to business data science to trade the academic track for fast-moving, applied environments. Today, I work in retail analytics, digging through transaction data and process flows to find where the happy path breaks down.

I tend to view everything - whether it is an organizational structure, a retail supply chain, or a data pipeline - as an information flow.

I try to bring a "machine-sympathetic" ethos to my work. Every system is optimized for something. In fraud analytics, that means asking "what does this system assume is true?" and hunting down where that assumption fails. In software, it means building tools that respect the underlying hardware through bounded memory, cache locality, and formats that stay out of your way.

• Languages: Rust (performance and correctness), Python (data/scripting, Polars-first), TypeScript (frontend), and Go.
• Infrastructure: Arch Linux everywhere. My homelab runs self-hosted Postgres, object storage, and Docker for reproducibility.
• Toolchain: Strong preference for tools that do their job and stay out of the way, like uv, bun, rustup, and pixi.