Madelon Hulsebos

Madelon Hulsebos is a faculty member at Centrum Wiskunde & Informatica (CWI), leading the Table Representation Learning (TRL) Lab and part of the Database Architectures group. She is also an ELLIS faculty member with the Amsterdam unit. Her research focuses on TRL, generative models for tabular data, and end-to-end systems for data analysis. She holds a PhD from the University of Amsterdam (2019) and conducted postdoctoral research at UC Berkeley (2020–2024). Her work has been supported by grants like the NWO AiNed Fellowship and the Accenture-BIDS Fellowship. She organizes TRL workshops at NeurIPS, ACL, and SIGMOD, and her research spans over 1.7M tables from GitHub (GitTables), schema analysis (SchemaPile), and dataset search systems. Key achievements include founding the TRL workshop series and leading the $1.5M DataLibra project to democratize structured data insights.

Education: PhD (2019) and postdoc at UC Berkeley. Grants: NWO AiNed ($1M+), NWO/AiNed DataLibra ($1.5M total). Affiliations: CWI, ELLIS Amsterdam, and collaborations with MIT, Sigma Computing, and industry sponsors.

Research Interests: TRL, generative AI for tables, data retrieval systems, and scalable data science. Recent Work: SIGMOD 2024 presentations, TRL@NeurIPS 2024 workshop leadership, and the SchemaPile corpus. Community Roles: ELLIS society member, co-organizer of DEEM workshop, and TRL Discord/Bluesky communities.