I build software that holds up in production. Right now that means agentic AI systems at Ford, and research on making medical AI evaluation honest.
Computer Science at Wayne State. Always glad to talk to people building something ambitious.
I build things end to end.
I'm a Computer Science student at Wayne State, and I've spent the last two years building software that real people and real organizations depend on. Ford brought me back on a return offer to help build the Fleet organization's first production agentic AI system. Before that I co-founded an emergency health platform, and worked as an engineer on a government-contracted system serving millions of requests.
Most of my work has landed in places where being wrong is expensive: healthcare, government systems, production data pipelines. That shapes how I build. I care about whether a thing actually works under load, whether the numbers behind it are honest, and whether the person on the other end is better off for it.
I also do research at Wayne State under Dr. Lihui Liu. My paper found that widely-used benchmarks for predicting ICU patient medications were leaking answers into training, which meant the scores the field was reporting were inflated. I built a corrected benchmark, a harder test set, and a model that beat five comparable approaches on it.
Where I've
made an impact.
Software Engineering Intern · Enterprise Technology
Building the Fleet organization's first production AI system, one that has to explain every decision it makes.
Undergraduate Research Assistant · Machine Learning
Found that a widely-used medical AI benchmark was leaking answers into training, and built a corrected one.
Co-Founder & Lead Engineer
Co-founded and shipped an emergency health platform, validated by the head of an ER department.
Software Engineer Intern
Made an $8M ARR government system meaningfully faster under real production traffic.
How I work.
I own the whole thing.
Code Emergency went from an idea to a HIPAA-compliant platform in production, and I was there for every layer: the schema, the API, the frontend, the deploy, and the conversations with the doctor who told us what actually mattered.
I work where being wrong is expensive.
Healthcare records. A government-contracted system at $8M ARR. Production pipelines at Ford. Those environments teach you to validate before you ship and to build the guardrails first, not after the incident.
I build with the people I'm building for.
The Head of Emergency Medicine validated Code Emergency, and what he told us changed the product, not just the pitch. At Ford I work alongside an AI engineer on a system whose decisions have to stay auditable by the people who depend on them.
Selected work.
eICU-KG
Medical AI models were being graded on a test that leaked the answers. I built an honest one, and a model that beat five others on it.
Mirage
A tool that makes AI agents testable. It intercepts what an agent tries to do and gives it a controlled, repeatable response.
Astra ERP
A business management platform built in a weekend at Google's Michigan DevFest. Finalist, and invited to present.
Code Emergency
A platform that gets your critical medical information to paramedics faster.
Let's build
something great.
Whether you're hiring, building something, or just want to talk through a problem, I'd like to hear from you. I answer everything.