Technical Product Leader · Seattle
Products where the details have to be right.
I've spent nearly 20 years building payments, cloud security, logistics and mobile products. I started as a software engineer, and I still like to be close to how things work. Lately I build practical AI, where rules handle the exact work and AI handles the messy leftovers.
Selected work
All projects →Hands-on AI prototypes. All data is synthetic, and all code is public.
Claims Pre-Check Agent & Rx Adjudication API
Flags pharmacy claim denial risks before submission. Rules check the exact fields; an AI agent reads the free-text notes.
Caught note-level denial risks that rule-only checks missed on a labeled eval set.
3-Way Reconciliation Data Generator
Generates ledger, processor and sponsor-bank data for cash advances and repayments, with realistic breaks injected.
Every break ships with an answer key, so recon automation can be scored.
Tool-Use Agent
A minimal Claude agent that decides when to call a Python tool, reads a local file and answers from it.
The request, execute, return loop behind every agent, in a few readable files.
Writing
All posts →Let rules do the exact work. Let AI handle the leftovers.
What building a claims pre-check agent taught me about where an LLM belongs in a regulated workflow.
Why three-way reconciliation is the long pole
Matching your ledger to a processor is tractable. Adding the sponsor bank is where weekends, delayed ACH and returns pile up.
Hiring, building, or want to compare notes? I'd like to hear from you.
Get in touch