AI Agents in the Enterprise - Trust, Governance & MCP Security l Dev in the details

Why Enterprise AI Agents Fail in Production: Trust, Risk, Observability + MCP Enterprises are moving beyond “LLMs for search” and into agentic workflows—LLMs stitched to tools that can take action. But once you move past the demo, reality hits: quality, security, governance, observability, and remediation become the real blockers. In this episode of Dev in the Details, Dev Rishi sits down with Travis Addair (Predibase co-founder & former Uber Deep Learning Infra lead) to unpack what’s actually stopping Global 2000 companies from scaling AI agents—and what the market needs to deliver next. We …