Andrew Milroy
Dominic Sartorio
Most enterprises can build a few AI agents. The economics change when they try to scale to hundreds. In this 30-minute analyst briefing, Andrew Milroy, Chief Analyst at Veqtor8, talks with Dominic Sartorio, Head of Product Marketing at Denodo, about the ROI of using an AI Data Layer for agentic AI. Andrew conducted an independent analysis of customers' enterprise-wide AI deployments and authored the study, showing a 290% average three-year ROI, and detailing how this is achieved.
Andrew explains that scaling and reuse are the key challenges an AI Data Layer solves; why integration followed by token consumption are the biggest cost drivers; and how the three common architectural patterns that enterprises run today (custom integration, lakehouses and hyperscaler ecosystems) hold up as agents multiply. He also shares what didn’t make it into the report, from infrastructure costs many large organizations haven’t yet thought through to the early warning signs that an architecture built for pilots won’t scale.
