WHAT: TWed Talk: "EFFECT: Human-AI System for Intelligence Activities"
WHO: Tiziana Hernandez & Ian Rudy
WHEN: 4p, Weds, 04 Feb 2026
WHERE: Winslow 1140
WEBEX: https://rensselaer.webex.com/meet/erickj4
EVENT PAGE: https://bit.ly/4k5EFTo
Please join us on Weds, 04 Feb (4p) as Tiziana Hernandez and Ian Rudy lead us in a discussion and demo of their fascinating work developing a RAG LLM chatbot that supports intelligence analyst question answering in shift handover settings. Pizza arrives approx. 3:30p, the talk begins at 4p.
DESCRIPTION: Intelligence analysts working in shift-handover environments must interpret, summarize, and query large volumes of heterogeneous reports under time pressure. In this talk, we introduce and demonstrate a retrieval-augmented generation (RAG) system developed over the past semester to support analyst question answering in handover settings. The system integrates document retrieval with a large language model to provide evidence-grounded responses to analyst queries over a shared corpus of reports.
The talk will include a live demonstration of the system, followed by a discussion of known system limitations and planned future improvements aimed at better supporting analyst workflows in shift-handover environments.
BIOS:
Tiziana Hernandez is a 2nd Year PhD student in Computer Science at RPI, and is advised by Dr. Tomek Strzalkowski. Her research interests lie in NLP and social computing, with a focus on online interactions, belief formation, and the analysis of information ecosystems and their threats- particularly in the era of generative AI. She is also interested in understanding and improving uncertainty estimation in LLMs through analysis of model outputs alone. Her broader research goal is to contribute to healthier, more trustworthy online information environments.
Ian Rudy is a 2nd Year PhD student in Computer Science at RPI, and is advised by Dr. Tomek Strzalkowski. His research interests include computer vision, visual accessibility, and otherwise visual/graphic domains in computer science. In addition to EFFECT, his current research involves physically-correct simulation of refractive errors (e.g. myopia/astigmatism) for analysis of web interface design accessibility. He has previously worked on research projects for computer vision systems for estimating the turbidity levels of water, and computer vision systems for monitoring oysters.