Join Us

We are always looking for motivated and talented people to join our group! We value diversity and inclusion, and we strongly encourage applications from underrepresented groups.

Open Positions

PhD Positions

We currently have two open PhD positions in the following areas:

  1. Causal Fairness in Sequential Decision-Making
    Work on developing causal notions of fairness for reinforcement learning and sequential decision-making systems. This position is funded through the DFG Priority Programme.

  2. Federated Learning with Formal Privacy Guarantees
    Develop new federated learning algorithms with rigorous differential privacy guarantees, with applications in healthcare and finance.

Requirements:

To apply, please send an email to niki.kilbertus@tum.de with the subject line “PhD Application”, including:


Postdoctoral Positions

We occasionally have postdoctoral openings. If you are interested in joining the group as a postdoc, please reach out directly with your CV and a short description of your research interests. Strong candidates may also be supported in applying for independent fellowships (e.g., Marie Skłodowska-Curie, Humboldt Fellowship).


Master’s Thesis

We regularly offer Master’s thesis projects for TUM students in the following areas:

Please send an email to niki.kilbertus@tum.de with the subject “Master’s Thesis Inquiry”, including your CV, transcript, and a brief description of which research area interests you most.


Internships and Visiting Researchers

We occasionally host research interns and visiting researchers. Please get in touch if you are interested in spending time with the group.


Research Projects

Our group works at the intersection of machine learning, causal inference, and statistics, with a focus on developing methods that are fair, transparent, and privacy-preserving.

Active Projects

Causal Fairness in Machine Learning

Algorithmic Fairness Causal Inference Decision-Making

We develop causal frameworks to rigorously define and enforce fairness in automated decision-making systems. This project investigates how causal assumptions can help distinguish between legitimate and discriminatory uses of sensitive attributes in predictive models.

Team: Prof. Dr. Niki Kilbertus, Alice Mueller

Funding: DFG Priority Programme SPP 1798

Related Publications

Privacy-Preserving Federated Learning

Federated Learning Differential Privacy Distributed Optimization

This project designs federated learning algorithms that enable collaborative model training across distributed data sources without sharing raw data. We provide formal differential privacy guarantees and study the privacy-utility trade-off in real-world applications.

Team: Prof. Dr. Niki Kilbertus, Bob Chen

Funding: BMBF

Related Publications

Uncertainty Quantification with Bayesian Causal Models

Bayesian Methods Causal Inference Healthcare AI

We combine Bayesian inference with causal graphical models to better quantify and communicate uncertainty in causal effect estimates, with applications in clinical decision support.

Team: Prof. Dr. Niki Kilbertus, Dr. Jane Smith, Clara Vogel

Funding: Helmholtz AI

Related Publications

Past Projects

Causal Discovery from Observational Data

Causal Discovery Structure Learning Completed 2022

Developed scalable algorithms for learning causal structure from large observational datasets, with theoretical guarantees under various faithfulness and Markov assumptions.

Team: Dr. Eva Fischer


What We Offer


Life in Munich

Munich offers an exceptional quality of life, combining a strong tech and research ecosystem with proximity to the Alps, excellent public transportation, and a lively cultural scene. The cost of living is high by German standards, but PhD and postdoc salaries at TUM are competitive.