One index for everything: Kasimir now connects knowledge from all your sources

The answer is in a document somewhere. The question is: which one? Company knowledge rarely lives in one place — it is scattered across cloud storage, team drives, local uploads, meeting notes and people's heads. Whoever needs something, searches. And searches.

This is exactly what our biggest update since launch addresses: Kasimir now brings the knowledge from all your sources together in one shared, searchable index — and answers questions with evidence and page numbers instead of guesses.

One index for every source

Connected cloud folders become part of Kasimir's knowledge — no matter where they live (be it Nextcloud, ownCloud, OneDrive, SharePoint or Google Drive). The same goes for uploaded documents and your notebook pages. Behind the scenes, Kasimir splits every document into meaningful sections, enriches them with context and makes them semantically searchable. The technique is called retrieval-augmented generation (RAG) — the AI doesn't answer from memory, it looks things up in your files.

You feel the result in every answer: Kasimir names the source, quotes the passage and, for PDFs, gives the page number — one click and you verify the statement in the original. For safety-critical topics like building codes or contract details, that evidence is the difference between a nice chatbot and a tool you trust.

Live or indexed — your choice per folder

Not all knowledge ages at the same speed. That's why connecting a folder offers two modes: Indexed builds a permanently searchable corpus that stays in sync automatically — ideal for policies, standards and project archives. Live reads the folder at question time, current to the second — ideal for the active project folder where new files land every hour.

Scans and plans can finally join in

A significant share of company knowledge exists only as scans: signed contracts, old building permits, photographed plans. Kasimir detects such documents automatically and reads them with vision OCR — including skewed or rotated pages. Afterwards they are as searchable as any other document, page references included.

Full transparency — for users and admins

Knowledge that flows in invisibly builds no trust. That's why the chat shows which knowledge sources would feed an answer — before you even ask. Every connected source opens in an explorer: browse folders, see files with their index status, search inside the source. And administrators get a central registry of all company sources — who connected them, how far indexing has progressed, where it's stuck.

And everything stays in Europe

The entire knowledge index runs on Kasimir's own infrastructure in Germany — from the embedding model to the vector database to OCR. Your documents never leave the EU for any of these features. That's not a feature, it's our foundation: data sovereignty is the reason Kasimir exists.

What's new at a glance

• Unified knowledge index across cloud folders, documents and notebook — with context enrichment for precise results

• Two modes per folder: permanently indexed or live to the second

• Citations with page numbers — every statement verifiable in the original

• Vision OCR for scanned PDFs, photos and plans — EU-hosted

• Source explorer: browse every source and search inside its content

• Live progress & admin registry: full visibility of all company knowledge sources

All features are available to every Kasimir customer today — no surcharge, no migration. Connect a folder and ask your first question.