REFEREE

REFEREE came from a simple frustration: delivering high-resolution video at scale is expensive, and the compute needed to produce it already exists in hundreds of millions of consumer GPUs sitting idle while people watch TV. The idea of shifting that work to the edge — literally the user's own desk — and letting RTX Tensor Cores and Radeon hardware do what they were built for felt like an obvious inversion of the standard streaming model. Choosing Rust for the backend was a deliberate decision for memory safety and execution speed; the service needed to feel invisible, and that meant no bloat, no leaks, and reliable teardown after every session. Tauri and Vite kept the GUI footprint minimal on the frontend side. The hardest engineering problem was not the upscaling itself — NVEncC and VCEEncC handle that brilliantly — but building a session model robust enough to manage per-user GPU pipelines with heartbeat-driven automatic cleanup, while keeping the REST API simple enough for any web developer to drop in.