QuantlysRecording an episode
- In the host console, start a Memory session and give it a project (your show).
- Send the link. The guest does a device check and joins from the browser.
- Turn captions on and press record. Both sides see the recording badge.
- After the call, build the Memory package: chapters, quotes, summary.
What you get per episode
| File | Format |
|---|---|
| Full video | One continuous HD 720p take (1280×720, 24fps). MP4 in Chrome, Edge, and Safari; WebM in Firefox |
| Captions | .vtt or .srt, for YouTube, Descript, or Premiere |
| Audio | Full episode as m4a (or audio.webm) |
| Clips | WAV chapter and quote cuts, when timed captions exist |
| Show notes | Markdown: summary, chapters, quotes, open threads |
No MOV and no 1080p or 4K. We only list what the recorder writes.
When to use something else
Quantlys records one composite take in the host’s browser, and remote guests come through as their network feed at up to 720p. Dedicated podcast tools such as Riverside or Zencastr record each guest locally on their own machine and upload separate tracks afterwards. If you need per-speaker tracks for a heavy edit, or guests on unreliable connections, use one of those.
Quantlys is for a different job: a conversation where you want the episode, the captions, and the written outline in one place with no extra transcription step. That works for interview shows, internal podcasts, and oral history.
Series, books, oral history
Sessions under one project roll up together, so you can reorder episodes and generate chapter headings across a series. The same flow covers a book drafted from conversations or an oral-history recording with a parent or elder.
Hosted or self-hosted
Use it at quantlys-meeting.com, or self-host with your own LiveKit, Deepgram, Supabase, and OpenAI/OpenRouter keys. The source is MIT on GitHub. For product meetings instead of episodes, see the AI meeting assistant that writes the PRD.