Research
I study causal inference and machine learning for the decisions entrepreneurs make.
Three pillars of my current research
- 1
Causality in Strategic Decision Making
Before a strategy is carried out, someone has to say why it should work. We write that reasoning down as a causal diagram, then use causal discovery and structural causal models to check it against data.
- Mapping theories of valuePreparing for submission
- Causal discovery in strategic management researchDRUID23
- Structural causal models in strategyAoM Proceedings, 2023
- 2
Machine Learning and Simulacra in Entrepreneurship
What can machine learning, and now language models, tell a founder? Earlier work predicted how ventures fare from their text, images and language. The newer work lets language models stand in for customers and give feedback before a product exists.
- Silicon sampling for product feedback in entrepreneurshipWorking paper, 2026
- Predicting outcomes in crowdfundingSmall Business Economics, 2020
- Learning to frameResubmit
- 3
Economic Complexity of Software
The software a country writes shows what it knows how to do. With Juhász, Wachs and Hidalgo, I measure how complex that code is and how it relates to a country’s development.
- The software complexity of nationsResearch Policy, 2026
- Creative complexityUnder review
- How researchers are using GitHub Innovation Graph data to reveal the “digital complexity” of nationsGitHub Blog, 2026
Collaboration
If you like to collaborate on research, especially with focus on machine learning and causality, feel free to reach out!
Summary
I study how data and signals shape strategic decisions under uncertainty, so that entrepreneurs and organizations can make better ones.
My work sits between strategy and entrepreneurship: how text, images and behavioral cues causally affect venture financing, competitive positioning and innovation outcomes. I use causal machine learning, natural language processing and computer vision on large observational datasets. My research with colleagues is cited by Porsche Consulting, Morgan Stanley, and in official reports of the Organisation for Economic Cooperation and Development (OECD) and the Publications Office of the European Union, particularly within the Joint Research Centre of the European Commission.
Methods
I mainly work with applied machine learning, especially natural language processing (NLP) and multimodal machine learning, which combines text, speech and video. More recently I added causal inference methods, including instrumental variables and causal discovery, to address endogeneity in observational data, and experimented with multi-agentic frameworks (simulacra). My PhD work taught me network analysis, graph theory and economic complexity metrics.
Highlights
- To the best of my knowledge, we had the first multimodal machine learning classifier for venture outcomes showcased in this publication.
- Together with my colleague Paul Hünermund, I co-founded the Causal Data Science Meeting in 2020. Since then, we had Guido Imbens, Susan Athey, Judea Pearl, among others as keynote speakers.
- One of the first peer-reviewed publications to measure the predictability of the Bitcoin price, based on social media emotions [tl;dr: highly volatile, reverse causality].
Reviewing and Service
- Reviewer for Academy of Management, ACM CHI Conference on Human Factors in Computing Systems (CHI), Computers in Human Behaviour, Journal of Business Venturing, Journal of Business Venturing Insights, Entrepreneurship Theory & Practice, Research Policy, Strategy Science, Journal of Management Studies.
- Web Coordinator, TIM Division (Academy of Management)
Grants and Collaborations
- 2025 Dutch Research Council – NWO Starter Grant (340,000 EUR) [Lottery]
- 2025 Collaborator: With Paul Hünermund (PI) – Independent Research Fund Denmark (5,296,599 DKK)
- 2023 Collaborator/Co-Supervisor. With Christian Seel (PI) – Dutch Research Council NWO M Grant (350,000 EUR)
- 2022 Project-Partner: With Christian Hopp (PI) – Swiss National Science Foundation (601,347 CHF)
- 2019 Google Research Cloud (5,000 EUR)
- 2018 Federal Ministry of Education and Research (112,000 EUR)
- 2018 Academics at Amazon (Workshop Cost)
- 2016 RWTH Project House ICT Grant, RWTH Aachen (96,000 EUR)
- 2014 PhD Fellowship, Dr. Werner Jackstädt Foundation (18,000 EUR)
- 2012 Research Fellow, Foundation for Economic Sciences (5,000 EUR)
Awards and Honors (Selection)
Poets & Quants Top 40 u. 40 MBA Professors, Santa Rosa, US; University-Wide Best Teacher Nominee (Top-5), Maastricht University, NL; Faculty-Wide Best Teacher Award, Maastricht University, NL; Borchers Badge, PhD Thesis Award, RWTH Aachen University, DE; Marshall Memorial Fellow, Marshall Fund of the United States, USA; Research Fellowship at the MIT Media Lab, USA; Fellow, Royal Society of Arts, London, UK Young Innovator of the Year 2013, Lab Finalist, Falling Walls, Berlin, DE; A.T. Kearney Scholar, Falling Walls, Berlin, DE; Rethinking Economics Fellowship, Foundation for Science, Frankfurt, DE; Awwwards Website Design Awards – Honorable Mention, USA; Participant at Y Combinator Startup School 2012, Stanford University, USA; Best Practice Paper Award, Stanford Medicine X, Stanford University, USA; Conference Grant by the Swiss National Science Foundation, Zurich, CH; Jörg Hogan Students Award (Dean’s List), Witten/Herdecke, DE; Dr. Ing. h.c. F. Porsche AG Talent Network, DE.