Built an automatic video captioning tool?

Hey Hacker News community!

I'm currently working on building an automatic video captioning tool using Flask, OpenAI, ffmpeg, and other technologies. The concept is simple: users can drag a video onto the website, and the site will automatically add subtitles directly into the video itself, rather than providing a separate text file.

I believe this could be a valuable tool, not just for personal use but potentially for integration with platforms like SORA in the future. However, I'm curious to hear from others in the community:

Have you ever used or attempted to build a similar tool? If so, what technologies did you use, and what challenges did you encounter? Do you have any suggestions or ideas for improving such a tool?

I'm particularly interested in hearing about any intelligent approaches or techniques that could enhance the functionality or efficiency of the video captioning process.

Looking forward to your insights and experiences!