I've built a career across very different worlds — and I bring all of them to the table. Range is my edge: I see problems from angles a single-track background can't, and I build AI that holds up in the places that matter.
I didn't take the standard route into engineering, and that's the point. Each field I've worked in left me with a different lens — and AI/ML is where they all come together. Here's what each one gave me.
My first world was one where decisions are made with incomplete information, under pressure, and people depend on you getting it right. It's where I learned to stay precise and calm when the cost of a mistake is real.
What it gave me: judgment under uncertainty, and empathy for the end user.Next I rotated through the departments of a military medical facility — the ER, the pharmacy, the COVID-19 vaccination team where I administered vaccinations, and quality assurance, where I audited clinics for protocol compliance and drafted the daily reports. I learned how good process prevents failure, and how to see a whole system rather than its pieces.
What it gave me: systems thinking, and a discipline for compliance and documentation I now bring to eval gates and audit trails.Then I taught myself to build the systems I'd only operated before. Engineering clicked fast — it rewards exactly the instincts my earlier worlds built: understand the problem precisely, design for edge cases, ship something you can stand behind.
What it gave me: the craft to turn ideas into working software.AI/ML is where all of it converges. I build systems with grounded retrieval, guardrails, and evaluation gates — drawing on a coordinator's view of the whole, an engineer's craft, and a caregiver's sense of who's on the other end of the output.
Where it leads: AI built by someone who's seen the problem from every side.Different worlds, one engineer. The range isn't a detour — it's how I see what others miss.
Different domains, same approach: understand the problem deeply, then build something that holds up in the real world.
Predicts patient-dropout risk in clinical trials on real registry data, with sponsor-level row-level isolation and a guardrailed RAG store — PR-AUC 0.70 on the AACT registry.
Classifies airline-support tickets urgent vs. normal, benchmarking RAG, plain LLM, an ML classifier, and zero-shot side by side to find the cost-optimal path for high-volume triage.
An AI copilot that triages Kubernetes GitHub issues — classifying them, answering via hybrid (E5 + lexical) RAG, and learning each maintainer's preferences across a multi-service stack.
Before moving into AI/ML, I was already building software for real clients — from Odoo customization and data support to municipal and banking applications.
If you're working on AI for a domain where the answer matters, I'd love to hear about it. Message me through the assistant in the corner, or find me here: