Speakers
Alina Budnyuk, Data Scientist and a Mathematician, SAS
Financial institutions depend on complex reporting processes built over decades, with critical business knowledge often buried in code, data pipelines and legacy systems. This session explores how agentic AI, powered by SAS Retrieval-Augmented Manager (RAM), can transform static code into a trusted enterprise knowledge layer. Using a BCBS239 regulatory reporting scenario, we demonstrate how banks can leverage lineage discovery, retrieval-augmented generation, large language models and agentic tooling to interrogate reporting logic using natural language, visualise dependencies, identify compliance risks and accelerate modernisation initiatives. The result is a shift from manual, document-driven processes to explainable, auditable and AI-assisted workflows, enabling faster insights while maintaining the trust, governance and transparency required in highly regulated environments.