This session was presented at BLUEPRINT 4D 2026.

How do we address the pressing and ever-growing challenge of managing and interrogating vast amounts of enterprise data, both within and outside PeopleSoft? Traditional methods often fall short in providing efficient and intuitive access to this data, leading to inefficiencies and missed insights. While PeopleSoft has made several strides in recent years utilizing OpenSearch to address this problem (Global Search, Real-Time Indexing, OpenSearch Dashboards, etc.), Generative AI has potential to unlock additional data insights. We will demonstrate how to leverage a platform-agnostic AI framework to interact with large language models (LLMs) within PeopleSoft to interrogate enterprise data. In this talk, we highlight a few specific use cases that showcase the practical applications of this technology: •We’ll explore a Retrieval-Augmented Generation (RAG) approach to interact with a knowledge base, facilitating a chat capability where users can ask questions about enterprise policies, procedures, and processes. •We’ll feed data directly from PeopleSoft into the AI, allowing users to securely ask questions on any data which they have access to. We rely on PeopleSoft Security and PeopleSoft Search Framework for dynamic OpenSearch queries, inheriting appropriate document level security to control what data the AI can access.


This session is eligible for the following CPE Credit: NASBA

Speaker: Chris Yeager, CACI Inc.