Access rights
Everyone only sees the documents they are allowed to see today.
Company knowledge with AI
Ask your documents, not your colleagues.
A company's knowledge sits in manuals, contracts, minutes and old emails. We make it available on request: you ask a question in plain language and get an answer from your own documents, with the source.
How it works
The technology behind it is called retrieval augmented generation, RAG for short. Instead of letting the AI answer from memory, the system first finds the relevant passages in your documents and gives them to the AI as a basis. The answer names the document it comes from. That way it can always be checked, and anything made up stands out.
Example: a service technician on the phone asks "Which torque applies to pump XY?" and gets the answer with a reference to the page and section in the maintenance manual.
Typical uses
- OnboardingNew employees find processes, templates and contacts on their own.
- Service and engineeringAnswers from manuals, data sheets and maintenance logs, also on the road.
- Contracts and quotesWhich terms apply to a customer, what does the framework agreement say about delivery times?
- Quality and standardsQuickly check what the process description or test instruction requires.
What we pay attention to
Up to date
New and changed documents are added automatically, outdated ones drop out.
Data protection
Depending on your data, with a cloud model or entirely on your own server, see Local AI.
Sources
Every answer points to where it was found. If the system finds nothing, it says so.
How we proceed
Initial conversation
Free and without obligation. We clarify documents, users and questions.
Review the inventory
Which repositories exist, in which formats, who may see what?
Pilot
One area, such as engineering or sales, with real questions from daily work.
Roll out
Further areas, training and support.
Costs
We set the exact scope in a clear quote after the initial conversation.
Knowledge assistant
from €1,000
One area with its documents, answers with sources.
With a local model
from €1,500
When the documents should not leave the building, excluding hardware.
Initial conversation
free
We look at your repositories and give an honest estimate of the effort.
Matching products
For local operation.
FAQ on company knowledge with AI
Which files can the system read?
The usual formats such as PDF, Word, Excel, PowerPoint, text and emails, including scanned documents with text recognition. Repositories such as network drives, SharePoint or cloud storage can be connected.
Can the AI make up answers?
The risk is much lower than with an AI without your own documents, because it relies on the passages found and names the source. It cannot be ruled out entirely, which is why the source is mandatory.
Are our documents used to train AI?
No. Your documents only serve as a reference for your own questions. With cloud models we choose providers and contracts that exclude training with your data, or run the model locally.
How much maintenance does it need?
Little, once the repository is connected. New documents are added automatically. At regular intervals we look at which questions remained open and improve the search.
What does a knowledge assistant cost?
A knowledge assistant for one area starts at €1,000, with a local model at €1,500 excluding hardware. With several repositories, rights management and integration into existing software it costs more; we set the scope after the initial conversation.
Further reading
From the blog (in German)
Terms explained
- RAG retrieval augmented generation
- Vector database how search finds meaning
- Hallucination when AI makes up answers
More AI services
AI solutions at a glance
Consulting, assistants, agents, company knowledge, infrastructure and training at a glance.
Learn more →Local AI
Openly available models on your own server or in an EU data centre. Setup from €1,500.
Learn more →AI agents
Agents that handle multi-step tasks, with your approval. Individual agents from €2,000.
Learn more →