Case Studies
Selected work from industry roles and my own ventures, labeled as such.
What this looks like in practice
The thread through my work: ML applied to the economics of expensive, constrained resources. Each story is labeled with the role I did it in, and claims are limited to what the public record supports.
NIL Valuation Platform
As founding engineer (contract) at ArcScore Inc.
ArcScore had a rough MVP and no production system: no pipeline, no infrastructure, and no defensible way to put a number on an athlete's NIL value.
I architected and built the platform end to end (a Next.js/React front end, a FastAPI back end, and a provider-agnostic Python data pipeline) and designed the core scoring and NIL-valuation methodology with two other modeling leads. I provisioned and now operate the production infrastructure: managed Postgres, a job queue, auth, secrets, CI/CD, and separate staging and production environments with observability on both.
Live production platform. I own significant portions of the application, data, infrastructure, and and engineering roadmap.
Capacity Planning Model
As an applied research intern at LinkedIn
The Applied Research team faced a confidential nine-figure hardware-allocation forecasting problem across LinkedIn's data centers.
I built a predictive model for it. The implementation is confidential, so the specifics stay with LinkedIn.
13.39% improvement over the existing methodology.
Accelerating Physics Simulations
As a scientific ML research intern at AMD Research
Physics-based simulations at AMD were expensive to run, and an ML surrogate is only useful if you can tell when to stop trusting it.
I worked on a hybrid method that interleaves physics-based simulation with neural network training and inference, and on Bayesian and ensemble uncertainty-quantification methods that flag unreliable surrogate predictions on the fly.
The method is public: I'm a named inventor on AMD's patent filing US 2025/0005236 A1 and a co-author of a NeurIPS 2022 ML4PS workshop paper. The filing describes the hybrid workflow completing the simulation in less time than the baseline. It states no speedup factor, so neither do I.
Infrastructure for Coaches
As founder of TechnicallyFit, my own company
A coach can only take on as many clients as they can program, check in on, and answer questions for by hand.
I built TechnicallyFit's platform so that work runs as software: a program engine, coach workspaces, and a cited sports-science knowledge base that the AI coach and outside AI assistants can query over MCP.
Infrastructure that lets coaches scale their businesses, and the foundation for TechnicallyFit's larger goal: the operating system for human performance.