Company Knowledge / RAG

Enterprise knowledge and RAG

Approved knowledge becomes discoverable, contextual and usable by AI.

Activate your knowledge View solution patterns

Document search at a white desk in the WIESCHER office with wooden windows and trees outside. — AI-generated scene based on our office premises.

Focus

The assignment

RAG connects a request with relevant approved knowledge sources without exposing all internal information indiscriminately.

What the solution covers

curated knowledge sources
role-based access
contextual delivery
source references and ownership

Suitable starting points

policies and manuals
product knowledge
recurring specialist questions
knowledge-intensive research

What people decide

Business teams own the sources, approvals and currency.

In brief and evaluation principles

What does RAG mean for enterprise knowledge?

Retrieval-augmented generation connects a request with relevant passages from approved knowledge sources. The model receives those passages as context for the request; this does not automatically make the content part of its training. RAG adds source context but does not guarantee a correct or complete answer.

What needs to be defined before operation

Sources need a defined purpose, accountable business owners, version, validity and a controlled process for approval, change and removal.

approved knowledge domains
role-based access enforced in the knowledge and retrieval system
traceable provenance and approval
handling of missing, conflicting or manipulated content

How quality is evaluated

Quality and coverage are evaluated with representative questions, expected evidence, difficult counter-examples and questions that should remain unanswered. Evaluation also covers source selection, permissions, conflict detection and expert reviewability. Consequential statements remain subject to review by a named business specialist.

Official primary sources

Content updated on 13 August 2026.

Editorially reviewed on 12 August 2026.

Retrieval-Augmented Generation (RAG) — CSRC Glossary (NIST AI 100-2e2025)National Institute of Standards and Technology (NIST)
Generative AI Models: Opportunities and Risks for Industry and AuthoritiesBundesamt für Sicherheit in der Informationstechnik (BSI)
Orientierungshilfe zu RAG-Systemen (Deutsch)Datenschutzkonferenz (DSK)
Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1)National Institute of Standards and Technology (NIST)

Which process should work better in your organisation?

We begin with the work, the people and the current process — then assess which form of AI genuinely makes sense.

Discuss your process