I currenlty work on Pruna AI which is a startup on efficient ML closely collaborating with the DAML research group (TUM). During my academic and industry work in ML, I realized that money, time, hardware, and environmental costs are citical barriers to apply ML in the real world. With the vision of making any AI model accessible for people and sustainable for the planet, I gathered the best team of cofounders to create Pruna AI. Now, Pruna AI develops technology that makes AI models significantly faster, smaller, cheaper, and greener in one line of code.

Beyond efficient ML, my research focuses on ML methods for uncertainty estimation for different data modalities, i.e. different input types (e.g. tabular, images, graphs, sequential, text data) and output types (e.g. categories, real values, positive counts). In particular, I am interested in defining desiderata, models and evaluation methods for practical uncertainty estimation in real-world applications. Other research interests include robust machine learning and structure learning like hierarchal and causal structures.

I did my Ph.D. in the Data Analytics and Machine Learning (DAML) group at the Technical University of Munich (TUM), where I was supervised by Stephan Günnemann. You can find my Ph.D. thesis on “Uncertainty Estimation for Independent and Non-Independent Data” here.

If you are interested in discussing with me, contact me per email or on twitter :)