Customer stories

IBB saves up to 120 hours a month in a single team. With one agent its own staff built on Langdock.

  • Emma Kudlich, Langdock
In short

The Institut für Berufliche Bildung (IBB) provides adult education and retraining in Germany. Its board backed AI from the start, and IBB trained employees early, from the fundamentals to the ethics of AI. Teams use Langdock in customer support, customer consultations and research. IBB's internal measurements show that one agent for standard documents saves a 15-person team 100 to 120 hours a month. The agent was built by an employee in the specialist department, not by the AI team.

  • 100–120 hsaved per month in a 15-person team with one agent
  • 3–4 minless processing time per support email, on average
  • 20 → a handfularticles read in detail per research topic, before and now

The starting point. Many inquiries, repetitive documents, a lot to read.

IBB's customer support handles a large number of customer inquiries every day. In customer consultations, qualified employees create the documents each consultation needs, a task that was time-consuming and repetitive.

In research and concept work, Natalie Güth, Team Lead AI & Data, used to work through every study and article in detail, around 20 per topic.

IBB's board wanted to use AI from the start, so the institute would not have to catch up later.

“In the past, I would have worked through every study and article in detail. Now I put them into Langdock and only check the facts from specific chapters in the relevant sources. That's a massive saving of many hours per topic.”

Natalie GüthTeam Lead AI & Data, IBB

The solution. Training first, then agents built where the work happens.

IBB started with an extensive training program, from the fundamentals of AI to its ethical aspects, to build a culture that is open to AI and asks critical questions about it. Employees were encouraged to try Langdock in their own work. When they had questions, they got help quickly from named contacts or from Langdock agents built for that purpose.

One agent per task, data pulled in through actions

In customer support, Langdock agents help process customer emails. In consultations, an agent helps create the required documents. Through actions, it retrieves data from an open web API and adds it to the document, which qualified employees complete with the agent in a few minutes.

“Our internal measurements have shown that in a team of 15 people, we can save 100-120 hours of work every month with just one agent for standard documents. What pleases me even more is that this agent was not dictated by us as the 'AI department,' but was independently created by an employee from the specialist department.”

Tim RöttgerHead of Digital Transformation, IBB

The outcome. 100 to 120 hours a month saved in one team, 3 to 4 minutes less per support email.

In one 15-person team, the standard-documents agent saves 100 to 120 hours a month, according to IBB's internal measurements. In customer support, processing time per email is down by 3 to 4 minutes on average.

Consultation documents take a few minutes, which leaves employees more time for conversations with customers. In research, Natalie Güth now reads a handful of articles in detail instead of 20, which leaves more room for creative and strategic work.

Langdock is a standard tool at IBB. Colleagues say, "Our friend Langdock will do that for us," and ask each other, "Have you tried that with Langdock?" where they used to ask, "Have you googled it yet?"

About IBB

The Institut für Berufliche Bildung (IBB, Institute for Vocational Education) provides adult education and retraining in Germany.

Industry
Public sector, education & non-profit
Team
Sales & customer service
Products
Agents
Emma KudlichBrand & Marketing, Langdock

Written January 2026