A selection of this year's partners

This year's course connected student teams with a wide range of organisations. A few examples:

• Anne Frank House
  After switching to 15-minute booking slots, the museum faced a new problem: visitors who simply didn't show up. Students used visitor flow and sales data to build predictive models for no-show rates, helping the museum sell the right number of tickets without turning people away, as well as spreading out the crowds within the museum to prevent congestion.
• Caeli
  Students tested and improved prediction models for tropospheric nitrogen dioxide values using meteorological datasets. Using a CatBoost approach, their results currently outperform existing published benchmarks.
• Reinland Water
  After visiting a treatment plant in Leiden, students tackled unpredictable water inflow: pumps delivering water in spikes rather than a steady stream, wasting energy and risking overflow. Their work explored how to equalise flow and improve resilience.
• Leiden Universitair Medisch Centrul
  How do you keep hospital wards below 80% occupancy while accounting for unplanned acute admissions? Students developed data-driven admission quotas at both tactical and operational levels to help the LUMC maintain flexibility without blocking capacity.
• Dutch Ministry of Defence / DefScholar
  Classified research is hard to navigate by design, but that creates real inefficiencies. Students built an AI-powered search engine to make defence-related research more accessible for personnel, working without access to the classified documents themselves.
• Sibi 
  After visiting Sibi's workplace, students built two predictive models for employee turnover using engagement signals and HR data. 

What students learned

What sets this course apart is that it doubles as a capstone thesis project, completed not individually, but in multidisciplinary groups. That combination shapes how students work and what they take away.

Across projects, students found that a real commissioner changed how they worked. Questions that would otherwise become assumptions could simply be asked. Clients brought problems they hadn't fully defined yet, and students had to help shape the question before they could answer it. Presenting to people outside the field raised the bar too: "People who see it have to actually understand it and not fall asleep."

It mattered, because it was real

What made it feel different from other assignments was simple: the work lands somewhere. Not in a drawer, but on the desk of someone who actually needs it. As one student put it: "This project was really educational. I wish I'd had something like this earlier in my bachelor's, or as an elective in a minor."

For some of the people involved, the connection might outlast the course. Students and organisations got to know each other through the work, and they might just end up continuing working together in the future. 

The thread running through most experiences: organisations get a fresh pair of eyes on problems they'd stopped seeing clearly. Students get something harder to manufacture in a classroom: figuring out what the data is actually worth, together with the people who need the answer.

More information about the project?

Bachelor Project: Business Case is coordinated by Sandjai Bhulai. Curious about how this course works, or interested in becoming a partner organisation? Reach out via s.bhulai@vu.nl