Dr. Paul von Bünau

Dr. Paul von Bünau

Managing Director

Dr. Paul von Bünau is the Managing Director of idalab. A mathematician by training, he has spent more than a decade supporting numerous biotech, pharma and medtech companies on AI strategy and technology. He lives in Berlin with his partner Louise East and two kids.

Paul is a frequent speaker on AI in healthcare, serves as a lecturer at the Europa-Universität Viadrina and loves to mentor nascent startups on what may, or may not, be possible.

Education

  • Ph.D. in Machine Learning with Computational Neuroscience (TU Berlin)
  • M.Sc. in Pure Mathematics (University of St Andrews, Scotland)
  • B.Sc. in Computer Science with Mathematics and Physics (University Potsdam)
Publications

  1. What I cannot simulate, I cannot assess: Understanding the Impact of Machine Learning Models for Alarm Reduction in Intensive Care Units

    Julian Beimes; Anne Rike Flint; Sophie A. I. Klopfenstein; Maximilian Markus Wunderlich; Mona Prendke; Amin Chaoui; Felix Balzer; Akira-Sebastian Poncette; Paul von Bünau Submitted 2026

  2. KI in der Medizin: Werden größere KI-Modelle auch praxistauglicher?

    Paul von Bünau; Louise von Stechow Tagesspiegel Background (Standpunkt) 2024 Read

  3. A standardized clinical data harmonization pipeline for scalable AI application deployment: Validation and usability study

    Elena Williams; Manuel Kienast; Evelyn Medawar; Janis Reinelt; Alberto Merola; Sophie Anne Ines Klopfenstein; Anne Rike Flint; Patrick Heeren; Akira-Sebastian Poncette; Felix Balzer; Julian Beimes; Paul von Bünau; Jonas Chromik; Bert Arnrich; Nico Scherf; Sebastian Niehaus JMIR Medical Informatics 11 2023 DOI

  4. Daten sind nicht das neue Öl

    Paul von Bünau; Sven Jungmann Tagesspiegel Background (Standpunkt) 2023 Read

  5. Predicting success of phase III trials in oncology

    Stephan Hegge; Markus Thunecke; Matthias Krings; Léonard Ruedin; Jan Saputra Müller; Paul von Bünau medRxiv 2020 DOI

  6. A mathematical model for the two-learners problem

    Jan Saputra Müller; Carmen Vidaurre; Martijn Schreuder; Frank C. Meinecke; Paul von Bünau; Klaus-Robert Müller Journal of Neural Engineering 14 (3), 036005 2017 DOI

  7. Explorative data analysis for changes in neural activity

    Duncan A. J. Blythe; Frank C. Meinecke; Paul von Bünau; Klaus-Robert Müller Journal of Neural Engineering 10 (2), 026018 2013 DOI

  8. An information geometrical view of stationary subspace analysis

    Motoaki Kawanabe; Wojciech Samek; Paul von Bünau; Frank C. Meinecke Artificial Neural Networks and Machine Learning — ICANN 2011, LNCS 6792, 397–404 2011 DOI

  9. RDRS and the Poor: Microfinance as Partnership

    Aldo Benini; Paul von Bünau; Mozammel Haque; Bhabatosh Nath RDRS Bangladesh and North Bengal Institute 2011 PDF

  10. Towards an unsupervised adaptation of LDA for Brain-Computer-Interfaces

    Carmen Vidaurre; Motoaki Kawanabe; Paul von Bünau; Benjamin Blankertz; Klaus-Robert Müller IEEE Transactions on Biomedical Engineering 2010 DOI

  11. Stationary subspace analysis as a generalized eigenvalue problem

    Satoshi Hara; Yoshinobu Kawahara; Takashi Washio; Paul von Bünau Neural Information Processing — ICONIP 2010, LNCS 6443, 422–429 2010 DOI

  12. Computing automorphisms of semigroups

    João Araújo; Paul von Bünau; James D. Mitchell; Max Neunhöffer Journal of Symbolic Computation 45 (3), 373–392 2010 DOI

  13. Finding stationary subspaces in multivariate time series

    Paul von Bünau; Frank C. Meinecke; Franz C. Király; Klaus-Robert Müller Physical Review Letters 103 (21) 2009 DOI

  14. Direct importance estimation for covariate shift adaptation

    Masashi Sugiyama; Taiji Suzuki; Shinichi Nakajima; Hisashi Kashima; Paul von Bünau; Motoaki Kawanabe Annals of the Institute of Statistical Mathematics 60 (4), 699–746 2008 DOI

Teaching

  1. Regulatory Compliance for AI in Healthcare: Four Case Studies

    Europa-Universität Viadrina Frankfurt (Oder)

  2. Foundations of AI Strategy: A Practical Perspective

    Europa-Universität Viadrina Frankfurt (Oder)

Focus

  1. Pharma Biotech

    Market Access and Pricing: Bespoke AI for Expert Teams

    Health technology market access has outgrown its tools. MA&P teams need bespoke AI fitted to their strategy, evidence base and workflows. We support from strategy to implementation.

  2. Investors

    Healthcare Buy & Build: AI Strategy and Implementation

    Leveraging data, workflows and patient/HCP touchpoints with AI can be significant value driver. As one team, we support from transaction to value creation plan to execution.

Articles

Buy, Build or Wait: AI for Market Access Teams

AI buy-vs-build is different for in MA&P. A walk through the trade-offs, and what they imply for the use cases everyone's talking about.

PICO Prediction: How good is off-the-shelf AI?

Using the first Oncology JCA (tovorafenib) as a test case, we asked Claude and Gemini for their predictions. Will they hold up?

The State of AI at World EPA Congress 2026

Notes from the World EPA Congress 2026 in Amsterdam, where the conversation has moved on from if to how. Success factors, lessons learned, impact.