Register for this event (HUID holders in person, members of the public on Zoom)

Greg Leppert is the Executive Director and Catherine Brobston is the Senior Partnerships Manager with the Institutional Data Initiative (IDI) at Harvard Law School Library. Founded in late 2024, IDI works with knowledge institutions -- from libraries and museums to cultural groups and government agencies -- to refine and publish their collections as data. One of IDI's key goals is to improve the quality and integrity of AI models by leveraging resources in the public domain, often stewarded by libraries. This work can also make library collections easier to find and search, enabling academics, public leaders, and beyond to do better research.

At this session of Data + Donuts, Leppert and Brobston will present IDI's latest work collaborating with libraries to turn collections into data for AI training. They'll discuss their collaboration with Boston Public Library to make thousands of historically significant government documents publicly accessible, and with Harvard Library to make nearly 1 million public domain books machine-readable.

This talk will be given in a hybrid format. Harvard affiliates are invited to register to attend in person, while members of the public are invited to attend online.

Since 2018, Data + Donuts has featured researchers and practitioners speaking about how they use data in their work, and on issues of data and society. We welcome those from the HKS community and beyond to share knowledge, discussion, and donuts. Explore the full list of past speakers. Video recordings are available upon request.

Anyone with a disability who would like to request accommodations or who has questions about physical access may contact library_research@hks.harvard.edu or call 617-495-7813 in advance of the program or visit.