Campus AI workflow
In developmentMatches researcher profiles against outside funding and collaboration opportunities, and explains why each one surfaced.
- PrototypeDelivery stageSources and ownership under review
- ExplainableDesign goalShow why each match appeared
- Cross-schoolIntended scopeWithin or across divisions
Accountability
How this use case is governed
- Service owner
- TritonAI solutions team and prospective research-service partners
- Human oversight
- Researchers and research administrators validate relevance and decide whether to act.
- Measurement plan
- Relevant opportunities surfaced, false matches, follow-through, and researcher usefulness.
- Data boundary
- Public description; any profile integration requires source-specific approval
Workflow elements
What the workflow uses
- Employee Activity Hub
- NIH RePORTER
- NSF Awards
- PubMed
- ORCID
- Semantic Scholar
How it works
How the workflow fits
The problem it addresses, what it actually does, and how far along it is.
Problem
Researchers and their support teams have to watch a lot of funding and publication sources that keep changing. Keyword matching either misses the relevant ones or buries them in noise.
Solution
The prototype explores how to parse the requirements of an opportunity and identify researchers whose publications, grants, and methods may be relevant. Candidate sources under evaluation include approved campus profile data and public academic sources such as NIH RePORTER, NSF Awards, PubMed, ORCID, and Semantic Scholar. The design goal is to explain why each suggestion appeared and link back to an authoritative source.
Current status
In development. We are still working out the data sources, consent, how often it refreshes, how to evaluate it, and who will own it before this becomes a pilot. A match is not an endorsement, and nobody should act on one without a researcher and a research administrator looking at it.