In a search
- Search across your own documents
- Ranked results with the passage shown
- Mix with web and Lens per query
Drop in files and Lens reads them into an Index you own. From then on it works like any other source you can search, chat over or hand to an agent, with the passage behind every citation and permissions carried through from where the document came from.
Three steps. Then the Index is available everywhere Lens answers.
Add PDF, DOCX, TXT or MD files today. Google Drive and Notion connectors are coming soon. Lens reads them into a private Index.
Turn the Index on for a search, a chat or an agent, alone or alongside web and Lens.
Answers come back grounded in your documents, with the source passage on every citation.
An Index is not a stale dump. It stays current and it stays inside your permissions.
We do not claim a compliance certification yet. What holds today: your documents stay in a private per-organization Index and are never used to train a model.
No. Documents in an Index answer your own queries. They are never used to train a model and never pooled with anyone else's data.
Only people you grant access. An Index is private to your org by default, and search, chat and agents see only the Indexes you point them at.