This is not a support role. We are looking for people who take initiative, learn by doing, and are comfortable sharing half-finished work with colleagues to get feedback. If you prefer to wait until something is perfect before showing it, this role is not for you.

What you will do

• Conduct a 2–3 week use case mapping exercise - interview colleagues, observe workflows, and identify the highest-potential AI opportunities
• Build and iterate on AI-assisted workflows using Claude and orchestration tools (n8n, Make, or similar) - targeting at least one deployed workflow within your first 6 weeks
• Present your progress, failures, and learnings at the bi-weekly cross-cohort showcase - visibility and knowledge sharing are a core part of your role
• Document successful workflows so that your colleagues can use them without your help - your goal is to make yourself redundant, then build the next thing
• Act as your department's first point of contact for AI tool questions, and advocate for what your department needs from the shared data and connector layer
• Collaborate with the Data & Connectors and IT & Security working students to access approved data sources and stay within security boundaries

What we are looking for

• Currently studying Business / Management, Computer Science, or Business Informatics - or demonstrable equivalent self-learning and project experience
• Has personally used AI tools and can give specific examples of how - not just aware that they exists
• Builds things in their spare time - side projects, personal automations, experiments. We want evidence of initiative, not just academic achievement
• Comfortable with ambiguity - the use cases you work on do not exist yet; you will define them
• Strong communicator - you will work with non-technical colleagues
• Basic scripting ability (Python or JavaScript) is helpful but not required - willingness to learn is essential

Interview requirement

Every candidate must build and present a 10-minute live AI agent demo as part of the interview process. The demo should address a realistic use case. It does not need to be polished or production-ready — we are assessing how you think about applying AI to real problems, not your slide design.