Transcribe a speech

By the Potok teamUpdated ~2 min read

To transcribe a speech, send the recording to the Potok bot on Telegram with timestamps on. Files up to 2 GB work. Export to DOCX so you can tidy the text, mark the best lines and pass it to the event team, then compare names and figures with the audio.

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

  • Clean text for a speech means trimmed hesitations and clear sentences, which is a decision you make while editing.
  • A DOCX file lets several people comment on the same copy, which is handy for event teams.
  • Timestamps give you the exact minute of the applause line or the key statistic.

How to do it

  1. Record from the room well. Place the recorder near the lectern or take a feed from the sound desk. Room echo hurts the result.
  2. Open the bot. Open @potok_now_bot in Telegram and start a chat with it.
  3. Switch on timestamps. They let you mark the quotable lines and find slides that match.
  4. Send the recording. Attach the audio or video as a file. For a long session recording, up to 2 GB, use the Upload button of the mini app.
  5. Export DOCX. Pick DOCX when the transcript is ready. A short result also arrives as text in the chat.
  6. Clean the text. Remove false starts and fillers, but keep the meaning and the speaker's voice. Check names, titles and figures.
  7. Share with the team. Send the DOCX for comments, then export PDF for the final copy that goes on your site.

What the result looks like

Example: a few lines from a keynote

[00:02] Speaker 1: Thank you all for coming. Let me begin with a question.

[00:09] Speaker 1: What would your town look like if every street had a garden?

[00:17] Speaker 1: Let us look at three places where this already works.

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
Over-polishing the speaker's wordsThe text no longer sounds like the person and may change the meaning.Cut fillers and repeats, but keep the wording that carries the point.
Ignoring slide contentThe text refers to the next slide with no explanation.Add short bracketed notes where the speaker points to a visual.
Forgetting permissionPublishing a full speech may need the speaker's approval.Ask before you post the text.
Sending a recording with audience noiseQuestions from the floor are mixed into the main voice.Check the places where more than one person spoke.

More

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

Frequently asked questions

Why export a speech to DOCX?

DOCX lets you edit, comment and track changes with colleagues. You can export PDF later as a fixed copy.

How do I get clean text?

Remove fillers and false starts while you listen back. Keep the speaker's meaning and any wording that makes a point.

Can I get the text of a talk from a video?

Yes. Send the video file, and the bot reads the audio track. SRT and VTT exports are also available for captions.

Does it label the speaker?

If you switch on speaker labels, voices appear as numbers. A single speaker needs only one label, and you can drop it.

How long can the recording be?

Files up to 2 GB are accepted. For larger ones, use the Upload button in the 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. Office Open XML (Wikipedia), Wikipedia · checked
  2. Transcription (linguistics) (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