Cloud Backup · AI Lens
Search 10 million files. Get the right one. In 3 seconds.
AI Lens runs on every backup — never on your endpoints. Find a person across ten years of photos, and search the words inside scanned PDFs and screenshots, even when the filenames tell you nothing.
What AI Lens finds
People, words and what's in the picture.
AI Lens runs automatically on files you've already backed up, turning storage into something you can actually search.
Faces & People
Tag a person once and BigMind finds every photo they appear in, across every backup and every device.
OCR & Text Search
The words inside PDFs, screenshots and scanned documents become searchable, not just the filename.
Image labels
Every photo is labelled automatically, so you can search "beach sunset" or "whiteboard" without tagging anything.
Tag a person once. Find them in every backup.
Point AI Lens at one photo of someone and it surfaces every picture they appear in, across every device and years of backups. Built for the photographer looking for a client, or the family looking for a grandparent.
- Trains on a single reference photo.
- 512-dimension face embeddings, matched with pgvector similarity search.
- Runs across every backed-up photo, old and new.
- Included from Standard up.
Search the words inside scans and screenshots.
PaddleOCR reads every PDF, screenshot and scanned document as it is backed up, so the text inside becomes searchable alongside file names, tags and folders — one search box across every device.
- Reads PDFs, screenshots and scanned documents.
- Finds a file by a word printed inside it, not just its name.
- Runs on the backup: nothing to install on your machines.
- Included from Standard up.
How it works
AI runs on the backup — not on your endpoint.
Runs on the backup
When a file lands in BigMind Resilience, the AI Worker pipeline picks it up. Nothing runs on your machines, so there is no CPU drain on your endpoints.
Models we host
SCRFD face detection with ArcFace embeddings, EfficientNet-Lite0 for labels and PaddleOCR for text: roughly 410 MB of ONNX models on our infrastructure.
Answers from the index
Faces, labels and text are written to the search index, so a search answers from the index instead of opening files one by one.
Where it shines
Three jobs AI Lens makes trivial.
The same engine, three very different workflows — each one a search that used to mean hours of manual scrolling.
Photographers
David Mojica finds client photos by name across more than ten years of shoots — no manual tagging.
Accounting & legal
Find the scanned contract, invoice or receipt by a word printed inside it, even when the file is called scan_0042.pdf.
Insurance
Pull every photo from a claim filing date even when the filenames are useless.
Pricing
Where AI Lens lives in pricing.
Faces, OCR and image labels are included on Standard, Pro and Enterprise, with nothing extra to buy. DR-Only protects machines for recovery rather than backing up files, so it has nothing for AI Lens to search.
| Tier | AI Lens (Faces, OCR, labels) |
|---|---|
| DR-Only | — |
| Standard | ✓ |
| Pro | ✓ |
| Enterprise | ✓ |
AI runs on every backup. So search just works.
Faces, text and what's in your photos — indexed from the data you're already protecting. Start free and point AI Lens at your own files.