Multi-Wing is a global engineering company with headquarters in Vedbæk, designing and manufacturing high‑performance fan solutions for customers worldwide. AI is a strategic focus area for us, and we are investing actively in building the next generation of digital tools and decision support for our engineers, operations, and commercial teams.
We are in an experimentation phase, trying things for the first time, learning fast, and building our future technology stack around Microsoft AI platforms. If you enjoy figuring things out, prototyping, and turning “what if…?” into working solutions, you will feel at home here.
About the role
We are looking for a Student AI Developer to help us design and build AI agents in our Microsoft environment, primarily using Copilot Studio, Foundry and Fabric – we are open to other suggestions if it fits into our stack.
You will join a small, focused AI team and work closely with business stakeholders to explore how agentic AI can support finance, engineering, sales, manufacturing and internal knowledge sharing. The role is hands-on and experimental: sometimes you will have clear requirements, other times you will help us discover what is possible.
If things go well, this role is intended to grow into a full-time AI developer position from January 2027.
You do not need to know everything on day one, but you should be excited to learn and build in these areas:
- Design and implement AI agents and copilots that solve concrete internal use cases (for example automation in finance processes, AI supported decision tools, product configurators, analysing and forecasting).
- Help set up and maintain data and knowledge integrations that agents can use, including document knowledge bases, structured data in Fabric and other internal data sources.
- Experiment with prompt design, system messages and tools/plugins to improve agent quality, reliability and user experience.
- Prototype multi-step and multi-agent workflows, such as routing requests, delegating sub-tasks or orchestrating calls to different systems.
- Work with the AI team and domain experts to translate business problems into agent tasks, evaluate results, and iterate based on feedback.
- Contribute to lightweight engineering practices: version control, documentation, small tests, and ensuring that what you build can be handed over and maintained.
You will be working primarily with:
- Microsoft Copilot Studio, Fabric, and Foundry
- LLMs and agentic patterns (tool use, RAG, multi-step workflows)
- Python, Power Platform and standard web tools where relevant
- Azure and the broader Microsoft ecosystem in collaboration with our IT function
You are most likely a master’s student in computer science, software engineering, data science, machine learning or a related field.
We imagine that you:
- Have some hands-on experience with programming in Python or a similar language, and are comfortable reading and writing code.
- Are genuinely curious about LLMs and agentic AI, and have tried building something yourself, for example with Azure OpenAI, OpenAI APIs, Copilot Studio, LangChain or similar tools (course projects, hackathons and private projects all count).
- Enjoy experimenting and learning by doing. You are ok with uncertainty, like to try things out, and can structure your findings so others can use them.
- Can break down a problem, design a simple solution, and iterate based on feedback from users and stakeholders.
- Communicate clearly in English, both spoken and written. Danish is a plus but not a requirement.
It is a plus if you:
- Have worked with Microsoft Azure, Power Platform, Copilot Studio, Fabric or Foundry.
- Have tried building chatbots, copilots, or internal tools for real users.
- Know the basics of RAG, vector search, embeddings or similar concepts.
