Documentation
Learn NeuroBridgeEDU.
A short, curated set of docs covering the things you'll actually want to know — how to install it, how it works, how to fix the common gotchas, and what's coming next.
Start here
The three docs that cover most questions.
Getting started
Getting started
Install the app, download a model on first launch, make your first recording — under three minutes from zero to first transcript.
Read
Concepts
How it works
The on-device pipeline from a microphone to a finished summary: Whisper for speech, Mistral or EuroLLM for summaries, all on your machine.
Read
Reference
Features
Recording, transcription, summarisation, search, exam-radar, and the rest of the surface — what each one does and when to reach for it.
Read
Help
Troubleshooting
Common problems and how to fix them, in the order people usually hit them. Recording, transcription, summarisation, the lot.
Read
All docs
Browse by category.
Six pages, one curated index. The full docs collection — all twenty-plus markdown files — lives on GitHub.
Quick answers
Common questions
If you're new and want to get oriented fast — start here.
Where do I start?
Read Getting started first — it walks you through install, first-launch model download, and your first recording. Most users are productive within ten minutes.
Is the app open source?
Yes — under our non-commercial open source licence. Source is on GitHub at EmminiX/NeuroBridge-EDU-OpenSource. Free for individuals and educational institutions; commercial licensing is a separate conversation.
Does it work offline?
Yes — fully. Once the AI models are downloaded on first launch (a one-time step), no internet is required to record, transcribe, or summarise. Cloud providers are optional and opt-in per session.
Where is my data stored?
Locally on your device, in a SQLite database in the OS application support directory. The full storage path and what's encrypted is documented in Privacy and data.
I'm stuck. Where do I get help?
Three places, in order of usefulness for most issues:
- Troubleshooting — covers the most common issues with the fixes.
- Support FAQ — shorter, more product-level questions.
- Open a GitHub issue or email us — see the contact page.
Which languages does it support?
Whisper supports 50+ languages for transcription. The summarisation models cover 35+ European languages well via EuroLLM, with Mistral particularly strong on French. The application UI itself is currently English-only — UI localisation is on the roadmap.
Looking for the source?
The full documentation lives on GitHub.
The pages here are a curated subset edited for the public marketing surface. The complete source — including internal architecture notes and per-feature documentation — is in the open-source repo.