Upload Session File

Supported: .vtt, .srt, .md, .txt, .mp3, .m4a, .wav
Auto-maps the dominant speaker to this name

How it works

  1. Upload your file → Stage 0 runs automatically (speaker detection)
  2. Watch progress in real time — you'll see the speakers found
  3. Confirm or rename the speakers on the same page
  4. Stages 1–5 run (cleaning, extraction, embedding) → vectors stored in Qdrant