UVC Partners: Data & AI - Master Thesis / Internship (f/m/x)
About us
UVC Partners is one of Europe’s leading early-stage venture capital firms. With our fourth fund generation of €250 million, we invest in disruptive B2B startups at the seed and Series A stage across Europe. Our focus areas include enterprise software, mobility, and industrial technologies.Headquartered in Munich and closely integrated with UnternehmerTUM, Europe’s largest entrepreneurship center, we are proud partners to some of Europe’s most ambitious founders - including Flixbus, Isar Aerospace, finn.auto, Proxima Fusion, and Aleph Alpha.
At UVC Partners, we’re driven by a deep passion for entrepreneurship and its power to positively shape society. Environmental and social responsibility are core to our DNA. We hold ourselves to the highest standards in everything we do. As a firm, we continuously reinvent ourselves to stay ahead and best serve our portfolio and investors. We believe in ownership over hierarchy and foster a culture built on collaboration, curiosity, and personal engagement. And while we’re serious about our mission, we also genuinely enjoy working together and building lasting relationships along the way.
Your mission
You will join a specialized team working at the intersection of applied research and venture capital, side by side with investors, AI researchers, and engineers building the backbone of our investment intelligence. This is a hands-on role with real stakes. You will be treated as a full member of the team, expected to take a research question from literature to a working prototype and to present your results to the people who make the investment decisions.Your profile
- You are enrolled in a Master’s program in Computer Science, Data Science, Data Engineering, Information Systems, or a similar program at a top university.
- You bring depth, or the ambition to build it, in at least one relevant toolkit, whether that is graph ML and graph databases (e.g. PyTorch Geometric), LLM tooling, tool use and evaluation, or probabilistic modeling (e.g. PyMC, NumPyro).
- You write clean, modular Python and have firm machine learning and statistics fundamentals.
- You are fluent with AI-assisted coding tools and you have judgment about them. You know what they do well, where they quietly fail, and you review everything they produce. You do not ship AI slop.
- You work independently. You have run a project where you set the direction yourself, and you can point to it.
- You are fluent in English, including technical language. It is what you will work, write, and present in. German is helpful but not required.
How we work
We are an AI-powered development team. Everyone here codes with AI assistance, and we treat that as a craft. You need to know which parts of a problem to hand over, where these tools fail quietly, and how to review output you did not write.
Agency matters a lot in our team. You get mentorship, context, and access to everyone. You do not get someone assigning you tasks every morning. Take a rough problem statement, come back with an approach, show drafts while they can still change, and flag what is not working. What you bring should outweigh what it costs us to support you.
Potential Research Directions
These are directions we could explore, with scope defined together at the start based on your background and where the work leads.
1. With Graph Neural Networks (GNNs) we would work on a graph-structured database, where people and companies are nodes and the relationships between them are part of the data, and ask what that structure can tell us that a table cannot.
2. With Program-aided Language Models (PAL) we would have the model write a small program for every number it pulls from decks and financial statements, so figures are accurately computed, because that is what makes it safe to point a language model at financial data.
3. And with Bayesian Updating we would keep our recorded understanding of a company moving as new evidence arrives, since the investment process is iterative and a view written down two years ago should not still be the view today. Your thesis could focus on one of these, though as a full-time member of the team you will work across several.
Academic Supervision: The Lead of the Data & AI Team is a PhD candidate at TUM's School of Computation, Information and Technology (CIT) and can supervise CIT masters students directly, so there is no separate supervisor search. However, you are free to choose your own supervisor and candidates from other universities bring their own.
What we offer
- Network: A desk at the Munich Urban Colab, our main office, with the option of a second one in Garching, plus access to the UnternehmerTUM ecosystem of startups, experts, and investors.
- Team & Culture: Monthly team lunches, our regular Stammtisch, free sports classes, Makerspace access for your personal projects, and regular team events.
- Compensation: An attractive salary for the full six months of your master thesis or internship.
- Ownership: Research that is yours end to end and feeds real fund decisions.
- Equipment: A MacBook or a Windows laptop, whichever you work best on.
- AI Tooling: Full access to the AI development stack we use daily, and a team that will teach you to use it properly.
- Real Data: Proprietary venture data that researchers rarely get access to, with the pipelines to use it productively.
Interested?
Please send a one-page CV, a current transcript, and a link to your GitHub or portfolio. Tell us which of these directions interests you most and why. We look forward to hearing from you!
Diversity is important to us
Diversity enriches our team. We explicitly encourage all qualified individuals to apply, regardless of gender, social and ethnic background, sexual orientation, religion, age, or disability.Contact
Unternehmertum Venture Capital Partners GmbH , c/o Munich Urban Colab, Freddie- Mercury-Straße 5, 80797 Munich, GermanyUVC Partners is an equal opportunity employer.
UVC Partners is an independent venture capital fund. For efficiency, we use certain shared services from UnternehmerTUM. This is why parts of the application process may appear UTUM-branded, and your application will be processed by UVC Partners and UnternehmerTUM.