Scriber
Status:
low-hanging-fruit| Maturity:production-ready| Last reviewed: 2026-03-04
What it does
Scriber is Sheffield’s own on-premise automatic transcription service, developed by Research Computing. It provides a web interface to open-source AI transcription models (based on OpenAI Whisper), allowing Sheffield researchers to transcribe audio recordings without sending data to external services.
Key capabilities
- Automatic transcription of audio and video recordings
- Runs entirely on Sheffield infrastructure — audio data does not leave the University
- Web-based interface — no installation required
- Backed by open-source Whisper models
- Free to all Sheffield staff and researchers
Fit for Sheffield researchers
Best-practice option for all Sheffield research involving audio transcription. The combination of free access, on-premise data handling, and Sheffield institutional support makes this the default recommendation for interview transcription, lecture capture, and qualitative research recordings.
Licensing and access
Free to all Sheffield staff and researchers. Access via Sheffield institutional login. Available immediately — no procurement or approval required.
Trust and governance notes
Data stays on Sheffield UK infrastructure. GDPR compliant for sensitive research audio, including interviews with research participants. Preferred over commercial transcription services (e.g. Sonix, Otter.ai) for any research involving personal data or confidential recordings.
Observed limitations
- Speaker diarisation (attributing words to specific speakers) is limited
- Transcripts require editing for research-quality outputs
- Service availability and capacity may be constrained for large-scale projects
- Dependent on Research Computing team for maintenance and updates
Related tools
Sources and evidence
- Scriber GitHub repository: https://github.com/rcgsheffield/scriber-service (accessed 2026-03-04)