Privacy

How we handle your videos and captions. Last updated September 2026.

What we process

When you upload a clip, we store it temporarily on our server disk to extract audio, run speech recognition, build caption timing, optionally romanize text, and (if you download) burn captions onto a new video file.

Third-party AI

Speech-to-text and romanization currently use Azure OpenAI / Cognitive Services (Whisper and chat models). Your audio/text is sent to those APIs solely to produce captions. We don’t sell your content.

Retention

Job files are ephemeral — meant for the session, not a permanent media library. By default we keep uploads and exported videos on the server for about 12 hours, then delete them automatically. Caption JSON and related work files follow the same cleanup. Don’t rely on us as backup storage; download your finished video while it’s still available.

Analytics & accounts

We don’t currently require accounts. Server logs may include job IDs, timestamps, and error messages for debugging. We don’t build advertising profiles from your videos.

Requests

For deletion requests or privacy questions, email buildwithdivyanshu@gmail.com with enough detail to find the job (time, filename if you remember it).