Tesco’s Warehouse Management System, built over four decades on mainframe COBOL, contains thousands of interconnected business rules that have evolved alongside retail operations. Before any modernization could begin, gaining a clear understanding of this legacy landscape was essential. In this session, you will learn how Tesco used agentic AI and retrieval-augmented generation to interpret legacy code, surface hidden dependencies, and reconstruct functional knowledge that had never been fully documented.

The focus was not on automating modernization, but on accelerating comprehension. By combining structured RAG pipelines with AI-driven agents, Tesco enabled engineering and product teams to build a validated, accurate understanding of how the existing system operates. The outcome was a reliable knowledge foundation that informed future architectural and strategic decisions for a mission-critical platform.

What You Will Learn

• How GenAI was applied to analyze legacy COBOL logic, batch jobs, and data flows in a retail warehouse management system

• How RAG pipelines were tailored for mainframe artifacts such as source code, job control language, and operational documentation

• How human-in-the-loop validation and governance ensured accuracy and trust in an enterprise environment

Who Should Attend

• Engineers, architects, and technical leaders working on legacy system discovery and understanding

• AI and RAG practitioners applying GenAI to complex enterprise codebases

• Product and delivery teams planning modernization initiatives

• Retail technology professionals interested in practical GenAI applications in retail systems