Key Takeaways
I built the first version of a project I personally needed — and I’m testing if it could be useful to others. Repo is public + I added a simple waitlist if you’d like to follow along.
Repo: http://github.com/Ga0512/video-analysis
Waitlist: https://iaap4qo6zs2.typeform.com/to/J43jclr2
What it does now:
- Process a video (file or URL)
- Split it into blocks for analysis
- Transcribe audio + caption frames
- Generate multimodal summaries (text + context)
Flexible setup:
- Run locally with open models (privacy, no API costs) Or connect your own API key (faster / larger models)
- Fully customizable: language, summary size (short/medium/long), persona, extra prompts
Ideas for future:
- Chat-with-video → ask questions directly about a video (using both frames + transcription)
- Export for AI parsing → structured export so you can feed the content into other AI workflows or databases
Possible pricing ideas:
- Pay-as-you-go credits for hosted usage
- Or a fixed subscription (X$/month) where you bring your own API key and just use the UI/UX layer
Why I’m here: Before polishing it into a MVP, I’d love some honest feedback:
Would you actually use a tool like this?
What do you value more: local mode (privacy, no cost) or API mode (speed, larger models)?
Does the chat-with-video/export direction make sense?
How would you prefer pricing?
If there’s enough interest, I’ll start building this in public (X) and share progress Thanks in advance
Not affiliated with Hacker News or Y Combinator. We simply enrich the public API with analytics.