Transcribe a brainstorm session

By the Potok teamUpdated ~2 min read

To transcribe a brainstorm session, send the recording to the Potok bot on Telegram, switch on speaker labels and export Markdown. Files up to 2 GB work. The result is searchable text, so a half-remembered idea from the middle of a loud hour takes seconds to find.

Transcribe a recording

Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.

Open @potok_now_bot in Telegram

Files up to 2 GB · export to TXT, MD, DOCX, PDF, SRT and VTT · speaker labels and timestamps · about 30 languages

Key facts

  • Brainstorms have many speakers who interrupt each other, so expect overlapping lines and mixed labels.
  • A keyword search jumps to the idea instead of replaying an hour of audio.
  • Markdown export keeps headings and lists, which suits notes you move into a wiki or a doc.

How to do it

  1. Place the recorder in the middle. Put the phone or recorder where it hears all seats about equally. A device in one pocket favours one voice.
  2. Ask people to finish sentences. Overlapping speech is the hardest part to read back. A quick rule helps.
  3. Open the bot. Open @potok_now_bot in Telegram when the session ends.
  4. Turn on speaker labels. Many voices are the point here, so labels save a lot of guessing later.
  5. Send the recording and export MD. Pick Markdown so you can paste the text under headings in your notes.
  6. Search for the ideas. Use your editor's search for words that mark an idea, such as maybe, what if, or we could.
  7. Credit and sort. Copy each idea to a list and add the right name. Group the list into themes.

What the result looks like

Example: a fast exchange in a brainstorm

[00:00] Speaker 1: Okay, ten minutes. How could we make the signup shorter?

[00:04] Speaker 2: What if we drop the password and send a link by email?

[00:08] Speaker 3: Or let people sign in with an account they already have.

[00:12] Speaker 1: Both go on the board. Next idea.

Sample text written by us to show the layout of a transcript. It is not the output of the recognition engine and not a real recording.

What Potok can do
FeatureDetails
Where it worksTelegram bot @potok_now_bot; large files go through the bot's mini app
File sizeUp to 2 GB
ExportTXT, MD, DOCX, PDF, SRT, VTT
Speakers and timeSpeaker labels and timestamps can be switched on
LanguagesAbout 30 languages with good published results; a few more with warnings
TranslationThe transcript is not translated into another language
DictationA separate Potok desktop app lets you dictate text into any program; it recognises Russian speech

Common mistakes

Common mistakes and how to fix them
MistakeWhat happensFix
Letting everyone talk at onceOverlapping speech comes out as broken lines and the best ideas get lost.Set a turn rule, or ask people to repeat key ideas one at a time.
Trusting the label to credit an ideaVoices that sound alike can swap numbers for a short stretch.Check the audio before you give someone credit for an idea.
Using one big file for a whole day of sessionsFinding a single session takes longer, and one failure costs you all of it.Send each session as its own recording.
Keeping the transcript but no list of ideasThe text is long and nobody reads it again.Pull the ideas into a short list within a day.

More

See the full list of languages, the tools and the glossary.

Frequently asked questions

Does the bot separate many speakers?

Speaker labels can be switched on and show numbers. With overlapping voices some lines may be assigned to the wrong number.

Why export Markdown?

Markdown keeps lists and headings as plain text, so it fits wikis and note apps.

Can I find a keyword in the transcript?

Yes, search the exported file in any editor. That is the main gain over audio.

What about people talking over each other?

Overlaps are the weakest point. Ask for turns when you can and check those lines by ear.

Is there a limit on session length?

Files up to 2 GB are accepted. Larger ones go through the Upload button of the bot's mini app.

Sources

Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.

  1. Speaker diarisation (Wikipedia), Wikipedia · checked
  2. Markdown (Wikipedia), Wikipedia · checked

Try it on your own recording

Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.

Open @potok_now_bot in Telegram