ZoneX
AI Sports Intelligence
A computer vision platform that transforms raw game film into real-time coaching intelligence. Built for the speed and stakes of NCAA athletics — where a single play adjustment can change a season.
The Problem
College coaching staffs are drowning in film. Every game produces hours of footage, every practice generates data, and every recruiting prospect has tape to review. The tools that exist are either too complex for coaching staffs to operate without dedicated analysts, or too simplistic to surface the insights that actually change outcomes.
Coaches need answers, not dashboards. They need to know which formation adjustment will exploit a defensive tendency, which player matchup creates an advantage, and they need to know it before the next drive — not after a data scientist spends three days with a Jupyter notebook.
The Approach
We built ZoneX as an explainable AI platform — not a black box that outputs numbers, but a system that shows coaches why it's recommending what it recommends. Every insight is traceable back to specific film sequences and data points.
- Computer vision models that automatically analyze raw game film — player tracking, formation recognition, route detection
- Unified intelligence layer that merges video, structured game data, and scouting reports into a single view
- Real-time play adjustment recommendations with explainable reasoning
- Designed for coaching staff usability — no data science expertise required
- Built to handle the scale and pace of SEC and Big 12 competition schedules
Architecture Decisions
The technical challenge was building a system fast enough for in-game use, accurate enough for coaching staff trust, and simple enough that a position coach can operate it between series.
- Real-time vision pipeline — optimized computer vision models that process film in minutes, not hours, with player and formation identification
- Explainable outputs — every recommendation links back to specific plays, tendencies, and film clips so coaches can verify the reasoning
- Data fusion architecture — unified data model that normalizes video, box scores, play-by-play, and scouting data into a single queryable layer
- Progressive complexity — surface-level insights are immediately accessible; deeper analysis is available for analysts who want to dig in
The Outcome
ZoneX reached programs in both the Southeastern Conference (SEC) and Big 12 Conference — two of the most competitive athletic conferences in NCAA Division I. The platform gives coaching staffs the ability to view, analyze, and adjust game plans with AI assistance in real time, turning film review from a multi-day process into a between-series workflow.
For a former Air Force Academy baseball player, building the tool he wished his coaching staff had is about as personal as a product gets.