Chifty Knowledge

Search OneDrive documents with AI

Chifty Knowledge connects OneDrive as a knowledge source. Your team asks questions in plain language and gets answers with source references – drawn from exactly the content that has been approved for it.

How does AI search OneDrive documents?

Chifty Knowledge connects OneDrive as a knowledge source. The AI searches only content approved for the person asking – the permission filter is applied on every request and cannot be bypassed by the AI. Answers appear in Chifty Chat and Chifty Office and cite the sources they are based on.

What gets connected

OneDrive is one of the data sources you can connect to Chifty Knowledge – alongside Google Drive, SharePoint, and HubSpot. Documents that already live in OneDrive become part of your shared knowledge base, without tediously copying or reorganizing content.

That knowledge is used where your team already works: in Chifty Chat, employees ask questions in plain language and get answers based on approved OneDrive content – with the sources used clearly referenced. In Chifty Office, the same knowledge flows into email reply drafts; every draft is reviewed and sent by you.

Permissions stay intact

In Chifty, knowledge is shared with the whole company, individual areas, roles, or specific people only – and the AI respects every one of these approvals, in every search, answer, and draft. The permission filter is applied at query time, directly in the search, before a result exists. The AI itself cannot bypass or influence it.

For your OneDrive content, that means: only what you have connected and approved as a knowledge source is used. Sensitive areas such as HR, finance, or management stay protected when classified accordingly – and every answer only cites sources the asking person is allowed to see themselves.

Honest limits

  • The connection needs read access to the OneDrive content that should serve as a knowledge source. Anything not connected and approved stays invisible to the AI – that is intentional, but it also means: you decide which content is usable, not the AI.
  • Chifty is a young product. We are continuously developing the connectors – if you have a specific use case, it's best to talk to us beforehand.
  • Answers are only as good as the connected content: outdated or contradictory documents show up in the answers. At least the source references make visible what an answer is based on.

Frequently asked questions

What about data protection when AI processes OneDrive documents?
The Chifty software is hosted in Germany, and AI processing runs in the EU by default. Transmission is encrypted with TLS throughout, and your content is not used to train AI models. With Google Cloud Vertex AI and OpenAI, we have contractually agreed that requests are not retained after processing. A DPA under Art. 28 GDPR is available on request.
Which Chifty product do I need for this?
Chifty Knowledge establishes the connection and forms the shared knowledge base. The knowledge is then used in Chifty Chat – for questions and answers with source references – and in Chifty Office, where it flows into email reply drafts.
Can the AI read all my OneDrive files?
No. The AI only uses content that has been connected as a knowledge source and approved for the person asking. The permission filter is applied on every request, directly in the search, before a result exists – the AI cannot bypass or influence it.
How can I search OneDrive with AI?
You connect OneDrive as a data source to Chifty Knowledge and decide which content is approved for whom. Employees then ask their questions in Chifty Chat in plain language – answers are based on the approved documents and cite the sources used.
Are my OneDrive documents used to train AI models?
No. Your content is not used to train the AI models we use. With Google Cloud Vertex AI and OpenAI, we have contractually agreed that requests are not retained after processing.