How TUP automates HR & Knowledge access with n8n and AI

Context

TUP GmbH & Co. KG is a software development company that builds fully tailored warehouse management systems. With around 160 employees and more than 40 years in business, the team is used to solving complex customer problems, but many internal processes still depended on manual coordination as the organization continued to grow in size and complexity.

In the last year, TUP created a dedicated AI and automation department led by Julian Stock. The team began using n8n as a practical way to introduce automation across the organization, starting with internal workflows where small improvements could remove daily friction quickly. From the beginning, n8n was positioned as a tool for real work, not experimentation.

Challenge

Before n8n, key internal tasks were handled through long chains of emails, calendar checks, and manual follow ups. Sick leave notifications and remote work requests required someone to read emails, interpret them, create calendar entries, and coordinate with administration — even though the company already operated with a high degree of technical sophistication in customer-facing systems.

TUP also faced the classic problem of internal knowledge being scattered across systems. Project information lived in Confluence spaces, but answering practical questions could still take too long, especially when someone needed specific details about a given project.

Finally, the company had an adoption challenge. As Julian explained, until recently AI and automation were not part of how most teams thought about solving problems. TUP needed a way to prove value fast, so that employees would start bringing automation ideas forward themselves. n8n became the vehicle for that shift.

Solution

TUP started by building internal automations in n8n that mirrored existing employee workflows, so adoption would feel natural. The work moved from "someone has to remember to do this" to "the process runs by itself."

One of the most valuable workflows was sick leave handling. Employees send an email, the workflow is evaluated automatically using AI, and n8n orchestrates the downstream steps — ensuring the right calendar entries and administrative communication happen without delay.

A similar approach was applied to remote work requests. With n8n, TUP checks company rules by looking at calendar usage, verifies whether a person has already used their allowed days, and then accepts or rejects the request accordingly.

For knowledge access, TUP built an internal chatbot by connecting OpenWebUI and its knowledge store using n8n. Employees can ask questions about specific projects and get answers without manually searching through Confluence spaces and pages.

Across these use cases, n8n became the standard orchestration layer. As the next step, the team is exploring how to use n8n and AI to automate documentation, so it can be generated alongside development rather than relying on developers to find extra time for it.

Results

The immediate impact was operational speed. Tasks that previously took ongoing attention from administration are now completed in seconds, freeing people to focus on higher-value work. As Julian noted, the goal was not to replace employees, but to remove work that was "really time-consuming and not really cool things" so they could reinvest that time into higher-impact work.

n8n also delivered measurable reliability and scale. In the last 30 days, TUP ran 1,282 executions with only 12 failures — a 0.94% failure rate — while saving 26 hours of work. That combination of volume, stability, and time saved helped n8n move from a promising tool to a trusted operational system.

Just as importantly, n8n created a cultural shift. Teams began bringing problems to Julian's group with a new expectation: if something is repetitive, it should be automated with n8n, not handled manually again.