Registry expertise, engineered forward
I'm a certified Oncology Data Specialist who abstracts across all cancer sites and supports registry operations for multiple facilities across several states. Every day I see where registry work loses time: re-reading long operative and radiation reports, chasing the same quality-control errors, and keeping up with coding manuals that change every year.
So I started building the tools I wished I had. Helix Core is my independent, self-built compound AI system: local language models, a curated knowledge base of registry rules and manuals, and specialized AI agents. It runs entirely on my own hardware, with a credentialed human reviewing every result.
I'm interested in registry leadership, and in helping the field adopt AI responsibly, with better, more machine-readable coding and notation standards.
What I bring
A snapshot of my experience. The full resume is available on request.
Certified abstraction
ODS-C credentialed, abstracting across all cancer sites with current SEER, STORE, SSDI, Grade and Solid Tumor rules.
Multi-facility operations
Supporting registry work for multiple facilities across several states, each with its own physicians, workflows and QC standards.
Quality & QC
Turning facility QC feedback into a documented error-pattern library, so recurring mistakes get caught before they happen.
Applied AI
Designs and runs local language models, custom agents and knowledge bases for operative notes, radiation oncology and dictation.
Privacy by design
Everything runs on local hardware. No patient data goes to cloud AI, and de-identification is built into every workflow.
Systems & automation
Self-hosted infrastructure, automated monitoring of coding-manual updates, and workflow tooling built to save abstracting time.
Helix Core
Local minds. Private data. One mission.
A compound AI system I designed and built for oncology data work: local language models, a registry knowledge base and specialized AI agents. The map below is live and updates itself daily.
Notes from the registry frontier
Articles on AI, data quality and the future of cancer registry work. Coming soon.
Why registries need local AI
Privacy, cost and control: the case for running models on your own hardware.
Notation that machines can read
Small changes to how we document that would make registry data far more usable.
Human-in-the-loop abstraction
Where AI helps, where it doesn't, and why the specialist stays in charge.
Request my resume
My full resume is shared on request. Tell me a little about who you are and why you're reaching out, and I'll reply personally.
Your details are used only to respond to this request.