ajay-desai9/PixelPitch

PixelPitch is a personal project with zero adoption signals (0 stars, 0 forks, no velocity). It assembles well-known, commodity computer vision techniques (YOLO, optical flow, K-Means) into a sports analytics application with no apparent user base or community. The README description suggests a working proof-of-concept, but lacks evidence of real deployment, user traction, or novel technical contribution. The architecture is straightforward application of standard CV libraries to a specific domain (football analysis). Platform domination risk is HIGH because Google (Cloud Video AI, MediaPipe), AWS (Rekognition), and Microsoft (Video Indexer) already offer turnkey sports video analysis. Established sports analytics firms (STATS Perform, Opta Sports, Hudl) have deep domain expertise and proprietary datasets this project cannot compete with. Market consolidation risk is MEDIUM because incumbent sports tech vendors could trivially add or license similar pipelines. Displacement is imminent (6 months) because the barrier to entry in this space is low—any well-funded player can assemble the same open components. The project has no moat: no dataset lock-in, no community, no novel algorithm, and no integration surface beyond academic interest. This reads as a portfolio piece or educational exercise, not a defensible product.