Close-up of the Chevrolet bowtie logo on a dark truck tailgate, with a taillight visible above.

Harvard Google Books Corpus

Role

User Experience Intern

Domain

AI Training
UX Research

Duration

12 weeks

Team

Kelly Ortega, Intern
Kai Frey, Intern
Meg McMahon, Manager

THE BRIEF

UXR @ Harvard Library

UXR @ Harvard Library

I'm conducting UX research on the Harvard-Google Books Corpus, a collection of 1 million digitized public domain books that were made available to researchers for training AI models, mining data, and building open-source public tools

I'm conducting UX research on the Harvard-Google Books Corpus, a collection of 1 million digitized public domain books that were made available to researchers for training AI models, mining data, and building open-source public tools

Harvard knows why users want to use the Corpus, but not what outcomes it has actually enabled. I'm doing a mixed-methods study using surveys and interviews with AI researchers and stakeholders to close this gap, delivering design recommendations to improve the Corpus for future AI research. This project is ongoing through late August 2026, but here's a snapshot of some details in the meantime

Harvard knows why users want to use the Corpus, but not what outcomes it has actually enabled. I'm doing a mixed-methods study using surveys and interviews with AI researchers and stakeholders to close this gap, delivering design recommendations to improve the Corpus for future AI research. This project is ongoing through late August 2026, but here's a snapshot of some details in the meantime

MORE PROJECTS

What else I'm working on

What else I'm working on

  1. Peer Full Text Search Usability Tests

  1. Peer Full Text Search Usability Tests

  1. Peer Full Text Search Usability Tests

I conducted 4 usability tests with Harvard students on the Library of Congress and HathiTrust's full text search feature to identify design opportunities to improve Harvard's own implementation of it

I conducted 4 usability tests with Harvard students on the Library of Congress and HathiTrust's full text search feature to identify design opportunities to improve Harvard's own implementation of it

  1. Trust in GenAI Chatbots Contextual Inquiry

  1. Trust in GenAI Chatbots Contextual Inquiry

  1. Trust in GenAI Chatbots Contextual Inquiry

I conducted 3 contextual inquiries with Harvard students into trust symbols they rely on using GenAI to help them find or understand information. What drives trust or distrust will shape how Harvard Library researches and designs its own chatbots

I conducted 3 contextual inquiries with Harvard students into trust symbols they rely on using GenAI to help them find or understand information. What drives trust or distrust will shape how Harvard Library researches and designs its own chatbots

Wireframe of a chatbot interface showing AI responses with trust indicator icons next to the assistant's messages.