You get a structured Markdown summary with a TL;DR, the key decisions made, a list of action items (each with an owner, deadline, and priority), open questions that need follow-up, and a short prep section for the next meeting. It's designed to drop straight into a Notion page, ClickUp ticket, or follow-up email.

Yes. The tool is built around the formats those products export — speaker-labelled transcripts with timestamps. You can paste straight from Gong, Granola, Otter, Fireflies, Tactiq, Zoom transcripts, or even a raw dump from a Google Meet recording, and it will normalise the output.

It comfortably handles a typical 30–60 minute call. For very long meetings (90+ minutes) the quality is best if you also paste a short meeting context line (date, attendees, project) so the model can anchor the action items. If you hit a limit, split the transcript into halves and run it twice.

Transcripts are sent to our AI provider to generate the summary and are not kept as searchable records on our servers. We don't share them outside the model provider. Even so, scrub sensitive client names or numbers from the transcript before pasting if your meetings touch regulated data.

Yes. You can paste a transcript in English or Bulgarian, and pick whether the output should match the transcript language, be in English, or be in Bulgarian. The action items, decisions, and section headings are all translated, not just the text.

You get a structured Markdown summary with a TL;DR, the key decisions made, a list of action items (each with an owner, deadline, and priority), open questions that need follow-up, and a short prep section for the next meeting. It's designed to drop straight into a Notion page, ClickUp ticket, or follow-up email.

Yes. The tool is built around the formats those products export — speaker-labelled transcripts with timestamps. You can paste straight from Gong, Granola, Otter, Fireflies, Tactiq, Zoom transcripts, or even a raw dump from a Google Meet recording, and it will normalise the output.

It comfortably handles a typical 30–60 minute call. For very long meetings (90+ minutes) the quality is best if you also paste a short meeting context line (date, attendees, project) so the model can anchor the action items. If you hit a limit, split the transcript into halves and run it twice.

Transcripts are sent to our AI provider to generate the summary and are not kept as searchable records on our servers. We don't share them outside the model provider. Even so, scrub sensitive client names or numbers from the transcript before pasting if your meetings touch regulated data.

Yes. You can paste a transcript in English or Bulgarian, and pick whether the output should match the transcript language, be in English, or be in Bulgarian. The action items, decisions, and section headings are all translated, not just the text.