How XIBIX Solutions cuts HR questions by 50% with AI workflows built on n8n.

Context

XIBIX Solutions GmbH is an IT services company in the Munich area with around 120 people. The team works across cloud development, cloud architecture, software development, and data engineering, supporting customers ranging from mid-sized organizations to large German enterprises in industries like automotive, finance, manufacturing and energy.

As the company expanded its internal AI journey, XIBIX needed a practical way to turn ideas into working workflows that both technical and non technical colleagues could actually use. Fabian Pagel, founder and acting CTO, wanted a platform that could help XIBIX move from isolated proofs of concept to repeatable internal services. For XIBIX, n8n became that platform, and n8n is now the place where the team is standardizing how they build and scale automations.

Challenge

XIBIX operates under strict information security expectations, influenced by certifications and the realities of enterprise customer work. Earlier experiments with tools like Make and Power Automate were useful, but they introduced tradeoffs around control, usability, and operating model. XIBIX specifically wanted to avoid a pure software as a service approach and instead run workflows under their own responsibility, which pushed them toward n8n and a self managed setup.

At the same time, HR was hitting a common scaling problem. XIBIX had a large set of HR documents and processes, including policies and onboarding guidelines. Employees needed quick answers like where to find vacation information or how to request HR paperwork. With a small HR team of about three to four people, repeated questions created a growing support load.

Finally, XIBIX needed adoption, not just technology. Fabian's goal was to make AI powered workflows easy to access inside familiar tools, so employees would actually use them. As he put it, "If I'm now in a chat window and I need to connect a tool, most of them don't use it." The solution had to feel simple, like chatting with an LLM, while still being governed and secure.

Solution

XIBIX started with the community edition of n8n to learn and validate the approach, then moved to the n8n Business plan to unlock capabilities needed for information security, including single sign on and automated onboarding and offboarding.

To prove value quickly and drive adoption, the team built what they call "Ask HR", an HR bot powered by n8n. Instead of building and maintaining a large custom infrastructure, XIBIX used n8n as the orchestration layer for a RAG style setup. In practice, n8n pulls HR content from Confluence, prepares embeddings, and uses a hosted vector database in Azure. n8n then runs the retriever and agent logic that answers employee questions based on the company's internal documentation.

A key step was distribution. XIBIX already runs an internal chat experience using Open WebUI. To connect that interface to n8n, the team built a lightweight API compatible with the OpenAI specification. That API inspects n8n workflows tagged for their internal "XGPT" and exposes them as selectable models. This means colleagues can create new agents in n8n, tag them, and make them available through a single entry point. Fabian summarized the value clearly: "It means I can now create hundreds of agents, and they are all available over this endpoint."

To make the experience even more accessible, XIBIX also began integrating the same n8n powered "Ask HR" experience into Microsoft Teams, meeting employees where they already work.

Results

With n8n in place, XIBIX reduced repetitive HR inquiries by more than 50%, and HR saves at least more than 50% of the time previously spent re-answering the same questions. That change freed the HR team to focus on higher value work instead of acting as a human search engine for internal documentation.

n8n also helped XIBIX broaden who can build. Fabian's co-founder Michael, who is not a software developer, built a workflow entirely on his own after seeing n8n — an experience Fabian described as "amazing", reinforcing that n8n could work across both technical and business profiles.

Beyond HR, the success of the n8n based bot sparked new internal demand. People started asking for n8n accounts and bringing forward ideas, including a tender evaluation workflow that would scrape tender data and use AI to support evaluation and document preparation. n8n is becoming a foundation for faster proofs of concept, less infrastructure overhead, and a repeatable way to operationalize AI workflows across the organization.