A Data Foundation for Enterprise Agentic AI: Key Requirements and Best Practices
Webinar Speaker: Donald Farmer, Research Fellow
Date: Thursday, June 25, 2026
Time: 9:00 a.m. PT / 12:00 p.m. ET
AI agents that reason, plan, and act across systems on behalf of users promise efficiency at an exponential scale. Yet TDWI research shows that while organizations are interested in agentic AI, many are still building the foundation needed to deploy it safely and effectively at enterprise scale.
Agents raise the stakes for AI governance and clean, trustworthy data. A chatbot that hallucinates may waste a user's time. An agent that hallucinates may book the wrong supplier, approve the wrong refund, or file the wrong claim, resulting in reputational harm, financial losses, or worse. It’s no surprise, then, that data quality and lack of business context are major roadblocks for organizations looking at agentic systems. Without trusted data, governed access, and reliable context, autonomous agents may amplify risk rather than productivity.
Drawing on Informatica and Google Cloud’s partnership and reference architecture, this session with TDWI research fellow Donald Farmer examines the key requirements that separate agent prototypes from implementations safe enough for production. We’ll explore the role of data management agents and Model Context Protocol (MCP) servers in agentic workflows to feed agents with trusted master data and provide metadata intelligence for rich semantic context. Key Takeaways:
- Why agentic AI raises the bar on data quality, since autonomous action compounds the cost of error
- How Model Context Protocol (MCP) servers can help ground agent reasoning in a trusted enterprise context
- What role policy-driven data access management plays when agents act across enterprise systems on a user's behalf
- Where to apply observability and lineage so agent decisions remain auditable in production
Date: June 25, 2026
Time: 9:00 am PT
Donald Farmer