Fix & Optimize Ai Video Processing – Silent Video to Auto-Explained Narration

I’m looking for a skilled developer or ML engineer to fix and optimize a video processing system that automatically generates explanatory audio from a silent video.
The system was functioning correctly initially, it detects the screen content of a video, builds a logical story from it, and then generates spoken narration to produce a complete, explained video. However, it’s now facing severe performance issues and stability problems.

❗️ Current Issues:
•  Processing has become extremely slow, even for short videos.
•  Sometimes the system auto logs out or crashes, halting the process entirely.
•  The slow and unstable experience is not acceptable for production or user-facing use.

✅ What You’ll Be Doing:
•  Diagnose and fix bottlenecks in the video analysis or narration generation pipeline.
•  Improve performance so even short videos process quickly and reliably.
•  Investigate and resolve the session timeout or crashing issues.
•  Ensure the entire system is stable and able to process videos end-to-end.

💻 Skills Needed:
•  Strong Python background
•  Experience with video processing libraries (e.g., OpenCV, ffmpeg)
•  Familiarity with TTS tools (e.g., ElevenLabs, Google TTS, pyttsx3, etc.)
•  Understanding of scene detection / object tracking
•  Ability to profile code performance and optimize memory usage
•  Optional: Background task handling (Celery, Redis), GPU acceleration

We’re ready to start immediately. If you deliver high-quality results, there may be potential for long-term collaboration on future AI-driven video tools.
Looking forward to your proposal!