Success Story: How BrühHaus Makes Experiential Knowledge Available for the Next Generation

Many mid-sized companies face the same challenge: experienced employees carry enormous domain expertise built up over years. They know typical failure patterns, repair approaches, edge cases, internal workflows, and small tricks that no manual fully documents.

As long as these people are around, much of this works through informal channels. A new employee asks an experienced colleague. A technician recalls a similar case. A solution turns up in an old document, an email, or a chat history.

But this model is becoming increasingly fragile. The skilled-worker shortage is intensifying the pressure. At the same time, experienced employees are retiring or reducing their hours — and the very knowledge that makes a difference day-to-day risks disappearing with them.

BrühHaus recognised this problem early and decided not to leave experiential knowledge to chance.

Working closely with BrühHaus, Hoch-AI developed the BrühAssistent: an AI knowledge assistant available directly inside Microsoft Teams.

The goal was not to replace employees with AI. Quite the opposite: the solution is designed to preserve, structure, and make accessible the knowledge of experienced team members.

Experienced colleagues can store knowledge in a central knowledge base. New employees can retrieve that knowledge later — simply, without lengthy searching or having to interrupt anyone.

The BrühAssistent answers questions in German, drawing on verified company documents from SharePoint. Teams is the central access point. SharePoint remains the source of truth.

The project established two central knowledge areas.

The first covers technical knowledge: repair guides, error-code lists, circuit diagrams, exploded-view drawings, and model-specific information. This knowledge is especially valuable for service and repair processes.

The second covers operational knowledge — for example, internal workflows and process information.

New or updated documents are maintained via SharePoint and indexed automatically. This means knowledge does not need to be copied into a new silo system. Employees continue working within the Microsoft 365 environment already in place at the company — lowering the barrier to adoption and making the solution practical for everyday use.

The decisive factor in knowledge management is not just storage. What matters is whether people can find the knowledge at the right moment.

That is why the BrühAssistent was integrated directly into Microsoft Teams — where employees already communicate.

Typical questions include: What does a specific error code mean for a particular model? How does the repair process with a loan device work? Is there a guide for a specific manufacturer? What steps are standard for a known issue?

The assistant searches the connected knowledge base and delivers an answer based on the available documents. Knowledge is not only stored — it is available inside the workflow.

An important part of the project is not just knowledge retrieval, but also making it easy to maintain the knowledge base.

Experienced employees can capture new knowledge directly in Microsoft Teams. The BrühAssistent provides a form card for this purpose, where experts can enter details such as manufacturer, model, error description, and solution in a structured way.

This is deliberately kept low-threshold. Nobody needs to know a separate portal or manually format documents. Knowledge is captured where the work already happens: in Teams.

After submission, the entry is stored in a structured format, added to the relevant knowledge base, and indexed automatically — so the new information is immediately available for future queries via the BrühAssistent.

This turns experiential knowledge into a reusable asset — not at some point in a large documentation project, but directly in the daily workflow.

The skilled-worker shortage is not just a recruiting problem. It is also a knowledge problem.

When new employees take longer to become productive, the burden on experienced colleagues increases. When critical information lives only in individual people's heads, dependencies emerge. When domain expertise disappears at retirement, it has to be laboriously rebuilt.

The BrühAssistent addresses exactly this. It makes experiential knowledge documentable, findable, and reusable — creating a knowledge store that grows with the company and is not tied to individual people.

For new employees, this means faster access to relevant knowledge. For experienced employees, it means fewer recurring questions. For the company, it means greater stability in critical processes.

The project was built on the Microsoft stack: Microsoft Teams, SharePoint, Azure, Azure OpenAI, and a Retrieval-Augmented Generation architecture.

Documents remain within the customer's Microsoft environment. The knowledge base is fed from SharePoint. The Teams agent provides the interface for employees.

The priority was not an AI demo effect, but a productive system that works in practice. The BrühAssistent is now live and running in production.

BrühHaus now has an AI-powered knowledge assistant that makes technical and operational knowledge centrally accessible.

The project demonstrates how AI creates concrete value in mid-sized companies — not as an abstract future topic, but as a practical tool against a very real problem.

Experienced employees can pass on their knowledge. Successive generations can access it. And the company builds, step by step, a knowledge base that lasts.

That is the real success of this project: AI is not deployed for its own sake, but as infrastructure for knowledge, continuity, and better onboarding.