These are the projects I recommend building instead of just buying something off the shelf.

This is an in-person hackathon in San Francisco for selected participants. Capacity is limited to 30 teams, teams must include 3–5 members, and there is no online or hybrid participation. Build ...

C++ inference library for multi-vendor edge NPUs. Current focus: OpenVINO 2025.x backend for Parakeet-TDT and Whisper models. Additional runtimes (Qualcomm QNN, AMD Ryzen AI Software) coming soon.