Audio & Speech services · 10 / 10
Audio & Speech
Verbatim transcription, speaker diarisation and sound-event labels.
We transcribe and segment speech to your style guide, verbatim or clean, with speaker turns, overlaps, non-speech events and timestamps aligned to the waveform.
Typical use cases
- ASR training data
- Call-centre analytics
- Voice assistants
- Sound-event detection
Sample deliverable
What the output looks like.
- 00:00.0SPK AThanks for calling. Can I start with your policy number?
- 00:04.3SPK BSure, it's, uh, N-W four four one two.
- 00:07.6A+B[overlap] Perfect… / Sorry, four one two.
- 00:08.2EVENT[keyboard typing]
- 00:10.4SPK AGot it. I'm pulling that up now.
Synthetic demonstration scene. No client data.
Variants
Transcription
Verbatim or clean-read, with timestamps at utterance or word level.
Diarisation
Speaker turns, overlaps and speaker attributes.
Sound events
Onset / offset tags for non-speech audio.
QA metric
Word error rate vs gold
≤ 5%
Diarisation error rate reported separately.
Export formats
- JSON
- SRT / VTT
- TextGrid
- RTTM
- CSV
How every project runs
- Scope call & sample review
- Written guideline & gold set
- Fixed-fee pilot with QA report
- Production in weekly batches
Next step
Scope your project with
an annotation lead.
A 30-minute call with the person who would run your project. We review a sample of your data, recommend an approach and send a fixed-price proposal within two business days.