About

I build high-reliability software systems where correctness, scale, and real-world constraints matter. I've been doing that professionally for 7 years, with more than 20 years of IT experience overall. I've worked with large enterprises like E. & J. Gallo Winery and startups like Martial Arts Technologies and Text2Drive, giving me a wide lens on how software gets built and shipped at very different scales.

Growing up in the Chicago area, my family ran one of the largest Camaro parts suppliers in the US, an early lesson in how technology and business operations intersect. That mindset stuck: my focus on customers and operations pushes me to engineer for business outcomes, not just system elegance.

I joined Tempus AI as an individual contributor and earned my way into the most urgent and challenging work the team had. My first significant test was diagnosing and resolving a circular dependency between the LIMS and OrderHub event-driven systems, a classic distributed systems problem with real clinical consequences. It required tracing race conditions across multiple message queues and implementing surgical fixes without disrupting production sample processing. That work recovered SLO compliance from 88% to 99% and opened the door to higher-stakes initiatives. Since then I've shipped nearly 1000 pull requests across 28 repositories, owned 13 epics end-to-end, and led the consolidation of multiple ordering intake mechanisms into a configurable microfrontend component system (270 PRs over two years), enabling every new product to launch 2–4 weeks faster while maintaining backward compatibility for a platform processing 36K+ monthly orders. I've also established faster software development strategies, including build tooling modernization that reduced hotfix deployment times from hours to minutes.

Right now I'm leading a greenfield NestJS microservice for mobile phlebotomy automation, designed from scratch using event-sourcing, multi-vendor webhook abstraction, and a high-durability inbox/outbox pattern that guarantees execution with resilience to transient errors and operational anomalies. I authored the complete infrastructure-as-code: GCP projects, Cloud SQL, Kubernetes workloads, AWS SES, Okta roles, and Sentry monitoring. The system enables Tempus to onboard new phlebotomy vendors without code changes and brings real-time appointment tracking visibility to operations staff for the first time.

The technical problems I'm most drawn to are the hard ones. High-concurrency systems that hold under load. Data integrity in regulated environments where corruption is unacceptable. APIs that scale without becoming a maintenance burden. What ties those problems together is that correctness isn't optional, and the architecture has to reflect that from the start.

That same discipline applies to the AI integrations I've created. I led the design and delivery of a human-in-the-loop document processing system for clinical requisitions at Tempus, where AI extracts patient and order data from uploaded PDFs and reviewers verify through attestation workflows before data flows downstream. Designing that system meant solving confidence thresholds, reviewer UX, audit lineage, and latency constraints simultaneously, under clinical-grade reliability expectations. That kind of system only holds together when the distributed infrastructure, UX, and ML assumptions are all designed in concert.

To formalize and deepen that foundation, I'm pursuing a Master of Science in Computer Science part-time, expected 2028. My focus areas include natural language processing and deep learning, both increasingly central to modern software systems and particularly relevant to healthcare applications like the work above. This academic grounding sharpens how I evaluate and utilize AI tooling, apply proven techniques, and keep my team anchored in facts rather than trends.

I bring an indispensable combination: deep engineering expertise, product thinking, and design intuition. Most engineers implement a spec. I can architect from scratch, and I can sit in a requirements conversation and simultaneously spot the UX edge case, the data model implication, and the downstream API impact. That cross-domain fluency means fewer handoff gaps, faster iteration, and systems that hold together end-to-end. I've led teams of up to five engineers across initiatives of significant scope, and I take that responsibility seriously — good technical leadership means the whole team ships better work, not just faster work.

My design background isn't decorative–it's analytical. My BFA in Visual Communication means I think about hierarchy, user perception, and interaction cost the same way I think about system reliability and API contracts: as things that either work precisely or quietly fail your users. That lens has helped me lead projects where the difference between good and world-class lived in the details no one wrote down.

Before moving into full-stack engineering, I cut my teeth doing UI development at Text2Drive, building real-time messaging widgets deployed across 200+ auto dealership sites. I also spent time freelancing, mostly on the PERN stack, where I learned to move fast, negotiate scope, and keep overhead low.

Outside of work: I play guitar and drums, volunteer as an ESL teacher through an international language exchange platform, and spend most of my free time with my family and pets.