We mark up your images and video, saying what each thing is and where, so AI can learn to recognise it.
Explore serviceData & annotation
Speech turned into text your voice products can learn from.
Real speech is messy: people talk over each other, there is background noise, and there are accents and specialist words. We write out and label audio to a clear style guide, marking who spoke and when, and review the work before delivery. The result reflects how people actually speak.
What we offer
Transcription (Audio Annotation) services.
- 01
Speech-to-text transcription
Verbatim or clean-read transcripts written to your style guide, including rules for fillers, false starts and numbers.
- 02
Speaker diarization
Speaker turns identified and labelled, including overlapping speech in calls, meetings and interviews.
- 03
Timestamps and segmentation
Utterance, phrase or word-level timestamps and segments aligned precisely to the audio.
- 04
Audio event tagging
Non-speech sounds such as music, alarms, laughter and background noise labelled with start and end times.
- 05
Multilingual and accent coverage
Transcription across languages, dialects and accents, including switching between languages within one recording.
- 06
Intent and emotion labelling
Caller intent, sentiment and emotional tone labelled for voice assistants and contact-centre analytics.
How we work
From guidelines to delivered data.
AI takes the first pass, trained people review it, and quality is checked before delivery.
- 01
Scope
We agree label definitions, edge cases, formats and quality targets with you.
- 02
Pilot
A small batch tests the guidelines before any volume work starts.
- 03
Annotate
Trained annotators label to the agreed guidelines, with tooling to match.
- 04
Review
Multi-stage quality review against the specification, with feedback loops.
- 05
Deliver
Data in your format, with notes on what was checked, then we iterate.
Where it helps
Typical projects.
The kinds of work this service usually supports.
Speech recognition training
Building and extending speech recognition datasets for new domains, accents or noisy environments.
Contact-centre analytics
Labelling calls for intent, outcome and sentiment so teams can understand why customers get in touch.
Voice assistants
Training and testing voice interfaces with realistic, well-labelled commands and conversations.
Media and subtitles
Producing timed transcripts and captions for podcasts, video libraries and training content.
Details
Output formats
- SRT
- WebVTT
- JSON with word-level timestamps
- RTTM (speaker diarization)
- Praat TextGrid
- CSV
- Plain text
- Custom schemas
Questions
Do you provide verbatim or clean transcripts?
Both. We agree a style guide with you covering fillers, repetitions, numbers, sound tags and unclear speech, so every transcriber follows the same rules.
Can you work with poor-quality audio?
Yes. We mark unclear or inaudible passages rather than guessing, and flag files that are too degraded to transcribe reliably.
How is sensitive audio handled?
We work under NDA if needed, restrict access to the people on your project and can redact personal details in transcripts on request.
More data & annotation services
All services
Match footage turned into data: who was where, what happened and when, for coaches, scouts and media teams.
Explore service
Software that watches video for you and flags what matters, from recorded footage to live alerts.
Explore service
Idea → Architecture → Prototype → Production → Scale
Talk to us about Transcription (Audio Annotation).
Tell us what you're trying to build, test, automate or improve. We'll tell you honestly where we can make a real difference, and where we can't.