What this tool does
You give it an audio file. It listens to the recording, writes down everything that was said, figures out who was speaking when, and saves a tidy written version into the company Knowledge Base — ready to read, search, and share.
Think of it like a diligent note-taker sitting in on your call. You hand them the recording afterwards; a few minutes later they hand back a typed-up transcript, neatly broken into “Speaker A said this, then Speaker B said that.” You don't have to scrub through the audio yourself.
The Transcriber app lives on the LIFE hub at hub.life-labs.dev/transcriber — there's also a Transcripts card on the hub home page. Finished transcripts are saved into the Knowledge Base reader.
Signing in & who can use it
The app sits behind the normal LIFE sign-in. When you open it, you'll be asked to sign in with your LIFE Google account if you aren't already — the same login you use for everything else on the hub. There's nothing extra to install.
There are two levels of access:
| If you have… | You can… |
|---|---|
| View access | See the list of transcription jobs and open finished transcripts. |
| Upload access | Everything above, plus upload audio to create new transcripts. |
If the upload area is missing or you see a message about access, you simply haven't been granted upload rights yet. Ask an admin to give you the transcriber.write permission and you'll be set.
Making a transcript
The whole thing is four steps: open the page, add your audio, wait a few minutes, open your transcript.
Open the Transcripts page
Go to hub.life-labs.dev/transcriber (or click the Transcripts card on the hub home page). You'll land on a page with an upload area at the top and a list of your jobs below it.
Add your audio
Find the box labelled “Drop audio files, or click to choose.” You have two options:
- Drag and drop one or more audio files straight onto the box, or
- Click the box to open a file picker and choose files from your computer.
You can add several files at once — each file becomes its own separate transcript. As soon as you add them, you'll see a little progress bar for each one while it uploads.
Watch it transcribe
Once a file finishes uploading it shows up in the “Your transcription jobs” list below, and the tool gets to work. The status updates on its own — you don't need to refresh the page.
- While a job is being transcribed, the list refreshes itself every few seconds.
- When everything is done, it slows down to a quiet check every 30 seconds.
- A short meeting usually takes a couple of minutes; longer recordings take a bit more.
| File | Status | Submitted | Result |
|---|---|---|---|
| team-standup.m4a | Published | Jun 18, 09:14 | View in KB |
| client-call.mp3 | Running | Jun 18, 09:21 | — |
| voice-memo.m4a | Queued | Jun 18, 09:22 | — |
Open your finished transcript
When a job's status turns to Published, a “View in KB” link appears next to it. Click it to open the finished transcript in the Knowledge Base reader, where you can read it, search it, and share the link with colleagues.
The transcript is named after the file you uploaded, so it's easy to find again later.
Understanding the status labels
Each job in the list shows one of these statuses. Here's what each one means in plain terms:
What your finished transcript looks like
The tool automatically tells the different voices apart and lays the conversation out by speaker turns — so it's easy to follow who said what, like reading a script.
The tool labels speakers generically — Speaker A, Speaker B, and so on — not by their real names. It reliably keeps the voices apart, but it can't recognise whose voice each one is, so it can't put real names to them. (The labels also aren't tied to a person across recordings — “Speaker A” in one transcript may be someone else in the next.) If you need names, add them yourself when you read the transcript.
Supported files & size limits
The tool accepts the common audio (and some video) formats your phone, computer, and meeting apps produce:
| What's accepted | Details |
|---|---|
| File types | m4a mp3 wav aac ogg flac webm mp4 |
| Maximum size | 250 MB per file. That's roughly several hours of normal audio — plenty for most meetings and interviews. |
| How many at once | As many as you like — each one becomes its own transcript. |
Phone voice memos are usually .m4a files — drop them straight in. Video files (.mp4, .webm) are fine too; the tool just listens to the audio inside them.
What happens to your audio
- Your audio is deleted after it's transcribed. Once the written transcript is created, the original audio file is removed — only the text transcript is kept.
- It's processed in Europe. Transcription runs on EU-based infrastructure (Dublin), in line with our data-handling standards.
- It's not used to train AI models. Your recordings stay private to LIFE and aren't fed into model training.
- Transcripts live in the Knowledge Base, behind the same LIFE sign-in as the rest of our internal tools.
As with any recording, make sure everyone on a call knows it's being recorded and transcribed. Treat transcripts as internal company material.
Tips for the best transcripts
- Clearer audio = better transcript. A decent microphone and a quiet room make a big difference; lots of background noise or crosstalk makes it harder to get every word right.
- One recording per conversation. Upload the whole meeting as a single file rather than chopping it up — the tool handles long recordings fine.
- Leave the tab open if you want live updates. The list keeps refreshing while the tab is open; if you close it, just come back later — your jobs and transcripts are still there.
- Always give it a quick read. Automatic transcription is very good but not perfect — skim the result for names, jargon, or numbers before relying on it.
Troubleshooting & FAQ
It says “Too big — the limit is 250 MB.”
Your file is over the size limit. Most meeting recordings fit comfortably, but very long or very high-quality files can be large. Try exporting the audio at a lower quality, or trim it, then upload again.
It says “Unsupported file type.”
The file isn't one of the accepted formats. Stick to m4a · mp3 · wav · aac · ogg · flac · webm · mp4. If you have something else (like a .mov), convert it to one of these first.
It says I'm “not signed in” or my session expired.
Your sign-in timed out. Just reload the page — you'll be signed back in via the hub, and you can retry the upload.
It says I “don't have access.”
You have view access but not upload access. Ask an admin to grant you the transcriber.write permission.
A job shows “Failed.” What now?
Usually this means the file was corrupted or unreadable. Hover over the red Failed label to see the specific reason, double-check the file plays normally on your computer, and try uploading it again. If it keeps failing, flag it to the team.
A job shows “Quarantined.”
The file was held for a quick safety review rather than published automatically — for example, it didn't pass the automatic file scan. This is uncommon. An admin can review and release it from the Knowledge Base.
Can I upload several recordings at once?
Yes. Drop in (or pick) as many files as you like — each one is transcribed separately and gets its own entry in the jobs list.
Why are speakers called “Speaker A” instead of their names?
The tool tells voices apart but can't recognise who each person is — so it labels them Speaker A, Speaker B, and so on. There's no way to have it learn or attach real names; if you need names, add them yourself when reading the transcript.
Where did my transcript go?
Every finished transcript is saved into the Knowledge Base, named after your file. Click “View in KB” next to the job, or browse the KB reader directly.
Do I need to keep the page open while it works?
No. The work happens on the server, not in your browser. You can close the tab and come back later — your jobs and finished transcripts will be waiting. Keeping the tab open just lets you watch progress update live.
Open the Transcriber app → and drop in your first recording.