Building Digital Services with AI: Vibe Coding for Early-Stage Innovation

Over one semester, you turn your own idea into a working web app, using AI coding tools. You do not need to be an experienced software developer.

You bring an idea from your studies, a side project, a club, or a problem you keep running into, and build it into a web app that runs: a frontend, a database, and a login, hosted so others can use it. Most of the code is written by the AI. Your work is to direct it, read what it produces, and find where it goes wrong. You begin in a no-code builder (such as v0, Lovable, or Figma Make for design-led ideas) and then move to a market-leading AI coding agent, such as Claude Code, Codex, or Cursor, for the full build. A hackathon in December, before the winter break, gives you a focused day to push the project forward.

“Vibe coding” here does not mean clicking until something appears. You work with context documents, a build log, and Git, and you learn to check the AI’s output rather than take it on trust. Two lecturers and two coaches support you throughout. At the end, you present your prototype and the idea behind it to a jury that combines technical perspectives (software and design) with entrepreneurial ones. The coaches help you prepare, and the showcase is a friendly, low-pressure event where you present what you built and what you learned.

Who it’s for

Master’s students across all programs of the KIT Department of Economics and Management (WiWi faculty). The programming basics from your bachelor’s are enough to start; you do not need to have built a full application or worked in a professional software development environment before. If you already build software, even better: you will get further and spend your time on the harder parts, a real backend, authentication, deployment, external services, and building small AI tools of your own.

Not enrolled in the WiWi faculty but think this is exactly the right course for you? Apply anyway, and we will see what is possible.

What you get

  • Your own web app, built and put online with guidance at every step. It is your own project and your own code, and the repository stays yours. There is no group work.
  • A cohort of ten students, with two lecturers and two coaches, and an open co-working slot each week for individual help.
  • No tool costs for you; we arrange access.
  • A final pitch to a jury with technical and entrepreneurial backgrounds, and for the strongest projects a chance to take the idea further within KIT’s start-up ecosystem.

The facts

  • 4.5 ECTS, about 135 hours. Around 40 of those are the sessions and the hackathon; the rest is your own project.
  • It can be credited towards your degree as the Teilleistung “Practical Seminar: Service Innovation” [T-WIWI-110887].
  • Graded, with no written exam. Half the grade is the final showcase (pitch, live demo, and what you learned); half is your build log and reflection. A project that runs into a dead end can still pass, because we assess how you worked and what you took from it, not only the finished result.
  • To pass, attend at least five of the six sessions, plus the hackathon and the showcase.
  • The course is an individual project. The lecture “Digital Services: Innovation & Business Models” [T-WIWI-112757] pairs well with it but is not required.

When

  • Wednesdays, 14:00 to 17:00: six content sessions, with open co-working in between at the TRIANGEL Open House.
  • Online kickoff on 21 October, first session on campus on 28 October.
  • A hackathon on 9 December, before the winter break, starting mid-afternoon with an optional open end in the evening.
  • Final showcase on 3 February at the TRIANGEL Studio.

How to apply

Apply through the WiWi seminar portal (https://portal.wiwi.kit.edu/ys/9374) by 19 July 2026 (end of day). Please include:

  • a few sentences on why you want to take part, and the everyday problem you would build for (you can still refine this until the course starts);
  • your program and current semester;
  • a short note on your experience with programming, building software, and AI tools (not required; it just helps us shape the content);
  • your transcript of records (Bachelor, and Master if available);
  • a short CV

Please also confirm that the schedule works for you: the Wednesday slot, the hackathon on 9 December, and the showcase on 3 February.

We keep to a fixed timeline: applications close on 19 July 2026 (end of day), you hear back from us by 20 July 2026 (end of day), and you confirm your place by 24 July 2026 (end of day).

Questions: daniel.heinz@kit.edu