SPORTS INTELLIGENCE / PRODUCT + DATA + MODELS
A multi-sport analytics platform built from the warehouse upward.
Founding engineering / full product ownership
Thousands of moving inputs reduced into an interface that lets someone make a decision.
Odds, schedules, lineups, player statistics, event data, weather and market feeds arrive with different identities, cadences and failure modes. The first product was the data layer itself: normalization, reconciliation, freshness rules, retries and a warehouse stable enough for every feature above it.
Under the dashboards sit simulation and pricing systems: seeded Monte Carlo game simulation, fair-odds projection, home-run scoring, EV and Kelly sizing, market comparison and backtesting. One simulation universe can price many markets while remaining reproducible from the same input snapshot.
The customer should not feel the complexity of the warehouse or model graph. The product turns it into research surfaces, ranked edges, shot and spray visualizations, player tools, bet tracking, AI insights and real-time alerts — with layered caching and read-time aggregation keeping the experience responsive.
The internal warehouse was isolated behind a dedicated read-only role and productized as a B2B developer API with authentication, usage metering, rate limiting, caching and OpenAPI documentation. The same infrastructure that powers Statyx can now become infrastructure for someone else.
Not a dashboard. A system that can keep turning changing external data into useful product state.
NEXT CASE / 02
iHateImagesYour image never left your device.