I’m building a desktop application in Python that can take an incoming camera feed and broadcast it live with minimal latency. The core of the job is the GUI, but every element of the interface must be wired to GPU-accelerated video processing so the stream stays smooth even at high resolutions and frame rates. Here’s what I need you to tackle: • Design and implement the desktop GUI (PyQt, PySide, or another modern toolkit you’re comfortable with). • Integrate GPU-based video pipelines—CUDA/OpenCL through OpenCV, or GStreamer with NVENC/VAAPI—so the heavy lifting never touches the CPU. • Build a simple “Go Live” flow: source selection, preview window, bitrate & resolution controls, and a clear status indicator while broadcasting. • Output the final stream to a standard RTMP or SRT endpoint so it can be picked up by common services such as YouTube Live, Twitch, or a private CDN. • Package everything into an easy-to-install desktop executable for Windows (cross-platform support is a bonus). Acceptance test: when I press “Go Live” on a 1080p60 feed, CPU usage should remain low while the GPU handles encoding, and the RTMP endpoint must show the stream with <1-second delay. If you’ve previously built GPU-accelerated streaming tools—or have strong OpenCV, FFmpeg, or GStreamer skills—let’s talk.