We have roughly twelve hours of meeting recordings a week. Nobody listens back; when someone asks about a decision the answer is "it is in the recording" and that is where it stays.
Turn meeting and voice recordings into summarised text
- 1. Asked
Community request opened
September 8, 2026 - 2. R&D
Speaker separation trial
September 22, 2026 - 3. Built
Summary and decision extraction
October 14, 2026 - 44. Live
Opening to everyone
HOW IT'S BUILT & HOW IT WORKS
Process summaryOur user brought a simple need: "Let me upload an hour-long recording and get back a summary split into decisions and to-dos, plus a transcript that shows who said what."
The recording is first split by speaker, then a speech recognition model tuned for Turkish transcribes every segment with timestamps. Once the transcript is done, a second step pulls out decisions, dates and assigned tasks and writes them up as a separate summary. Output is SRT, TXT or a shareable page.
weekly-ops-2026-10-13.m4a[09:14:02] UPLOAD: weekly-ops-2026-10-13.m4a received (58:12).
[09:14:40] DIARIZE: 4 speakers separated, 312 segments.
[09:15:31] SUMMARY: 6 decisions, 9 to-dos extracted.
[09:15:34] DELIVER: Summary page and SRT ready (HTTP 200).
Drop in a Zoom, Meet or phone recording; m4a, mp3, wav and mp4 are accepted.
Speakers are separated, every sentence is timestamped, and you match the names once.
Decisions, dates and assigned tasks arrive in a short list, apart from the transcript.
THREAD & BUILD LOG
4 RepliesOur subtitle tool already produces a timestamped transcript. Two things are missing: telling speakers apart and pulling decisions out of the transcript. We are trying both on three of your recordings first.
Our meetings run half in Turkish, half in English. If the transcript does not break when the language switches, we would use it from day one.
In progress. Speaker separation and mixed language work; we are now on the summary step, which lists decisions and assigned tasks separately. We are aiming to open it to everyone at the end of October.
SIMILAR SOLUTIONS
Transcribe videos once, subtitle them in every language
One upload gives a transcript with speakers separated and SRT / VTT subtitles in every language you publish in, timings intact.
Alert me when one value on the page changes
The moment a price, a stock level or an exchange rate changes on any shop or market page, it fires a Telegram or Slack alert.
Redact personal data from a PDF for real
Permanently redact sensitive personal data in PDFs hundreds of pages long, with WebAssembly inside your browser, without sending them to any outside server.
GOT ANOTHER PROBLEM?
Write down the idea on your mind, let the community vote, and we build the most requested ones within a few days.
Turn meeting and voice recordings into summarised text
Transcribe #063 · In sprint 2, aimed at the end of October
- 01
Asked
September 8, 2026
Community request opened - 02
R&D
September 22, 2026
Speaker separation trial - 03
Built
October 14, 2026
Summary and decision extraction - 04
Live
Opening to everyone
https://app.incapon.com/transcribe/uploads/weekly-ops
weekly-ops-2026-10-13.m4aHOW DOES IT WORK?
Speaker separation, Turkish-tuned speech recognition and decision extraction are chained into one job.
Upload the file
Drop in a Zoom, Meet or phone recording; m4a, mp3, wav and mp4 are accepted.
Who said what
Speakers are separated, every sentence is timestamped, and you match the names once.
Decisions set apart
Decisions, dates and assigned tasks arrive in a short list, apart from the transcript.
Good to Know
Follow it and it lands in your dashboard as a ready template on launch day; upload your first recording straight away.
DISCUSSION LOG
We have roughly twelve hours of meeting recordings a week. Nobody listens back; when someone asks about a decision the answer is "it is in the recording" and that is where it stays.
Our subtitle tool already produces a timestamped transcript. Two things are missing: telling speakers apart and pulling decisions out of the transcript. We are trying both on three of your recordings first.
Our meetings run half in Turkish, half in English. If the transcript does not break when the language switches, we would use it from day one.
In progress. Speaker separation and mixed language work; we are now on the summary step, which lists decisions and assigned tasks separately. We are aiming to open it to everyone at the end of October.
SIMILAR SOLUTIONS
Transcribe videos once, subtitle them in every language
One upload gives a transcript with speakers separated and SRT / VTT subtitles in every language you publish in, timings intact.
Alert me when one value on the page changes
The moment a price, a stock level or an exchange rate changes on any shop or market page, it fires a Telegram or Slack alert.
Redact personal data from a PDF for real
Permanently redact sensitive personal data in PDFs hundreds of pages long, with WebAssembly inside your browser, without sending them to any outside server.
GOT ANOTHER PROBLEM?
Send us the dull digital chores you still do by hand. Let's solve them together with the community.