Zum Hauptinhalt springen

Quicklinks

Winter Semester 2026/27 · Open to everyone at UoC· 9 ECTS via the "Deep Learning Project" module · Application required

Solving real-world, industry-relevant agentic AI use cases

The first IBM watsonx Agentic AI Challenge at the University of Cologne brings together the university, IBM, and industry partners to work on real-world Agentic AI challenges.

Following the success of the IBM watsonx challenge in Switzerland, the program is now launching in Germany for the first time at the University of Cologne.

Participants form teams and develop solutions for enterprise-specific AI use cases provided by partner companies, using IBM watsonx and other AI technologies. The challenge combines theoretical and practical sessions with hands-on project work and direct interaction with industry partners.

The challenge takes place during the Winter Semester 2026/27.

Participation and Credits

The challenge is open to participants from across the University of Cologne.

Bachelor’s and Master’s students who would like to receive academic credits (9 ECTS) can participate as part of the Deep Learning Project module provided it is available as part of your study program. Participation without receiving credits is also possible.

Participation requires completing a Course Application survey before the start of the challenge. The information provided in the survey will help us form balanced, interdisciplinary teams and, if demand exceeds the available places, determine the final participant list.

If you are interested in participating, please complete the Course Application using the link below. The link was corrected on 11.09.2026.

https://survey.uni-koeln.de/index.php/536455?lang=en

Deadline is 30.09.2026. We may consider late submissions with reduced priority. 

What to Expect

The challenge combines lectures, practical labs, and project work with industry partners. Planned topics may include:

  • Introduction to Agentic AI for Business – understanding AI agents, enterprise applications, opportunities, limitations, and trustworthy AI
  • Agentic Systems – how agents work, tool use, memory, orchestration, and single- versus multi-agent approaches
  • Prompt Engineering, RAG & Tool Calling – techniques for working effectively with large language models, external knowledge, and tools
  • Agent Development & Best Practices – developing, testing, evaluating, and improving agentic applications
  • Design Thinking & User Stories – translating real business challenges into well-defined AI use cases
  • Solution Presentation & Storytelling – communicating prototypes, business value, and results effectively

Hands-on labs are expected to cover topics such as working with IBM watsonx, prompt engineering, retrieval-augmented generation (RAG), agent development, and selected open-source AI tools.

The exact lecture and lab program may evolve depending on the participating industry projects and the technologies most relevant to the teams.

Industry Partners

Students will work on real-world challenges together with participating companies. Industry partners provide relevant AI use cases and interact with the student teams throughout the project, providing domain expertise and feedback.

Current industry partners include:

  • REWE digital
  • Barmenia.Gothaer

Further industry partners will be announced.

Team

The challenge is organized jointly by the University of Cologne and IBM, with contributions from academic, transfer, project-management, and technical experts.

  • Prof. Dr. Aleksandar Bojchevski — Computer Science, University of Cologne
  • Dr. Dadhichi Paretkar — Transfer & Spin-off Coach, University of Cologne
  • Matthias Fuhrmann — IBM Cologne Location Manager
  • Tatiana Hiller — Project Manager, IBM
  • Mathias Werwie — Technical Lead, IBM

Additional IBM experts, university colleagues, and representatives from the participating industry partners will contribute to lectures, labs, mentoring, and project feedback.

Benefits

  • A unique real-world experience with cutting-edge AI and enterprise technology
  • Hands-on exposure to practical AI use cases in business
  • Insights from IBM experts on Agentic AI technologies and methods
  • Direct collaboration with industry partners
  • Experience developing an AI solution from use case to prototype
  • Networking opportunities with companies, IBM, and participants from across the University of Cologne

Course Module Description

This is a practical, semester-long capstone course in which small student teams develop an end-to-end machine learning or deep learning solution to a real-world problem. Teams work iteratively with regular supervision and feedback, guided by intermediate milestones. The course concludes with a final delivery, typically including a prototype, a presentation, and a report. Assessment considers the quality of the solution and methodology, as well as teamwork, progress, and communication.