| What it does | An API that verifies each RAG response against its sources
| Anti-cheat for online assessments
|
|---|
| Stage | Not given
| Not given
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | FR
| Not given
|
|---|
| Pricing model | Contact for pricing
| Not given
|
|---|
| Pricing | Not given
| Not given
|
|---|
| Key features | - Response verification against source documents
- Confidence levels: certain, probable, or I don't know
- Nightly automatic improvements with proof
- REST API with coming Python and TypeScript SDKs
- LangChain and LlamaIndex integration support
- MCP server integration option
- Data hosted in Paris, Europe
- Full verification in French and English
- Graceful degradation if API fails
- User data never used for AI training
| - Tab-focus, copy/paste, fullscreen-exit and single-face checks via webcam
- Phone observer for second angle detection
- Desktop lockdown with kiosk mode and keyboard hook
- On-device processing with MediaPipe classifiers
- Evidence control options: none, thumbnails, or short clips
- Server-authoritative verdict system
- Webhook notifications for threshold crossings
- GDPR, FERPA and BIPA compliance by design
|
|---|
| What makes it different | - Honest "I don't know" instead of hallucinated answers
- Proven improvements: 31 out of 32 correct on sealed test set
- Drops in after your existing RAG: one API call only
| - Runs entirely on-device; video never leaves the device unless explicitly captured
- No ML training or camera pipeline required to get started
- Three flexible assurance tiers for different exam stakes
- Privacy-first design with explicit consent and evidence minimization
- Integration in under five minutes with one API call
|
| Community votes | 0 | 0 |
| DR (Domain Rating by Ahrefs) | Not checked yet ● no change since last check | Not checked yet ● no change since last check |
| Trust Flow (Majestic) | 0 | 0 |
| Citation Flow (Majestic) | 0 | 0 |
| Referring domains (Majestic) | 0 | 0 |