Hello friend, I'm Cody.
TypeScript · JavaScript · React · Next.js · Node.js · Python · FastAPI · C# / .NET · SQL · Postgres · Git · Docker · Vercel · OpenAI · Claude · Claude Code
🏍️🏎️ Trackday Tuner
A mobile race engineer for riders and drivers. Log a session in under 60 seconds, compare setup changes against rider feedback, and keep conditions, outcomes, and history in one place. Its retrieval-based AI guidance uses your actual track data instead of generic setup advice. Built with Next.js, TypeScript, Supabase, and Tailwind CSS.
Check-in and live timing for a sim racing venue. Drivers scan a QR code at their rig, a front of store TV shows live "fastest tonight" standings, drivers follow their own laps on their phone, and staff run the floor from one dashboard. A .NET agent on each simulator ships laps through a durable, idempotent outbox, so a network drop or restart never loses a lap, and the database enforces one open assignment per rig instead of trusting the app to.
🏍️💬 Attend
A service advisor may be managing 40 customer conversations at once—sales questions, service updates, parts arrivals, and customers who have gone quiet. CTX Connect uses ambient AI to summarize each thread, surface the next action, and prioritize the inbox by what actually needs attention instead of what arrived last. It combines shared texting, follow-up tasks, team assignments, and SMS compliance in one workflow. Built with Next.js, TypeScript, Twilio, Prisma, and Neon Postgres.
Subscription VOD platform for martial arts training. Structured courses, entitlement logic, Stripe billing, Mux video delivery, and admin publishing workflows.
🧬 OncoPath
Live app that turns public ClinicalTrials.gov records into plain language explanations for patients. Built around an evaluation harness that caught bugs in both the generator and the automated judge, then got checked against human review before any score was trusted.
🚲⚡ Wattsmith
Local first cycling workout builder with FTP based power targets and verified .mrc / .erg export. Exported files are parsed back and diffed against the source automatically, so a bad file never reaches a real training ride. 119 tests and counting.
🚲💨 RideSense
Cycling analytics app that dedupes TrainerRoad, Strava, and uploaded ride data into one canonical training timeline. Content hash deduplication, encrypted OAuth token handling, deterministic analytics, and grounded Q&A without double counted TSS or hand wavy AI coaching.



