This session was presented at BLUEPRINT 4D 2026.

ERPs such as JD Edwards and PeopleSoft produce vast amounts of operational data, yet architectural constraints and siloed access limit its usability by modern AI and agentic frameworks. This gap prevents organizations from moving beyond static reporting to AI-driven insights and intelligent, automated decision-making. This session explores how bringing AI directly to enterprise data—leveraging Oracle AI Data Platform and OCI Generative AI models —can unlock the full value of ERP systems. Using modern AI techniques such as vector databases, retrieval-augmented generation (RAG), and large language models (LLMs), organizations can enable secure, natural-language access to ERP data while preserving existing JD Edwards and PeopleSoft investments. The presentation will walk through practical reference architectures built on OCI, including the use of the Oracle 26 AI Lakehouse to unify structured ERP data and unstructured enterprise content for AI-driven analytics and automation. Real-world use cases will demonstrate AI-assisted ERP reporting, contextual knowledge discovery, and NLP-enhanced workflows that reduce manual effort and accelerate decision-making. In addition, the session will outline best practices for enterprise-grade security, data governance, model lifecycle management, and the design of AI agents using agentic frameworks to orchestrate multi-step ERP interactions. The session will also offer a forward-looking view into emerging patterns such as agent-to-agent collaboration, highlighting how coordinated, multi-agent workflows can further extend ERP automation while keeping data close to its source. It will also cover scalable deployment patterns on OCI. Attendees will leave with actionable architectural patterns and implementation strategies to deliver AI-led business outcomes using their existing Oracle ERP data assets.


This session is eligible for the following CPE Credit: NASBA, IIBA

Speaker: Amarnath Aitha, Oracle America: Inc