A repository of past, present, and future work that the AI Lab has made and is making possible.

Active Projects 

  • Applying an LLM and computer vision to create alternative text for accessible photos and images.

  • Artificial Intelligence (AI) Guide for Humans- This guide is designed to help cultivate decision-making and critical curiosity around generative AI. Developed by multiple librarians, this guide aims to be a good starting point to thinking critically about how to use generative AI. Includes open-access and in-house resources to help facilitate thinking and discussion.

  • Metadata extraction using a vision capable LLM- International Snow Science Workshop (ISSW) has their conference papers hosted by MSU. Every two years it is a massive workload to extract all the metadata from these papers, so we are exploring using open-source vision LLMs to extract this information from the ISSW papers. 2026 Conference papers had their metadata extracted using Qwen-3.8 on CatChat.
  • LLM and Vector Empowered Search- As search functionality continues to change with new LLM and vector features we are exploring how best search can be implemented across the Library's collections and information.

 

Past Projects

  • Retrieval Augmented Generation (RAG) using Open-Source AI- Using CatChat, a workflow was develeped to take PDFs and turn them into Markdown and then add them into a knowledge base that utilizes RAG search.
  • Prompting Copilot to perform data cleaning and analysis on Excel documents- We developed an example of how to take raw, unorganized data along with a detailed document of definitions and desired outputs to create prompts that use Copilot to clean and analyze the data.

 

Research

  • Responsible AI: Tools for values-driven AI in libraries and archives: Project website and News release. Viewfinder was developed by librarians at Montana State University, University of Montana, James Madison University, and Iowa State University, based on research conducted between 2022 and 2025, made possible in part by funding from the Institute of Museum and Library Services. 
  • Modeling Transparency: Designing Model Cards for Responsible Generative AI Assignments- 2 MSU Library faculty were awarded Montana University System Teaching Scholar grants to work with 8 MSU faculty to develop discipline-specific AI teaching tools, which have been published and are freely available.