Vandalizer
Overview
Developed by AI Engineers and Data Scientists at the University of Idaho, Vandalizer is a document processing platform which leverages powerful AI models to extract and format data from technical documents. It allows research administrators to chain extraction, prompting, and formatting tasks into workflows to make them more efficient than ever before. Workflows developed in Vandalizer are flexible, the results they create are accurate and reproducible, and the platform itself is private and secure.
Vandalizer Release Roadmap
Vandalizer is in the testing and development phase with University of Idaho and Southern Utah University acting as first users. Based on input from the Research Administration (RA) community, our formed community of practice, and feedback from RA colleagues, the AI4RA team is looking for a way to allow other entities to gain access to the tool without comprising our workplan or security protocols. We intend to release more information once we have a formal plan in place.
Vandalizer Open Source Released!
We are excited to announce that Vandalizer is available to the public in two new editions. Vandalizer is an AI-powered document processing platform designed to help research administrators automate repetitive tasks like extracting information from funding announcements, summarizing complex documents, and generating structured outputs from unstructured files. By building reusable workflows, teams can process research documents faster and more consistently while maintaining human oversight.