The messy truth of your AI strategies

A podcast conversation with Hema Raghavan, co-founder of Kumo.ai, covering the real challenges companies face when implementing AI at scale. Topics include shadow AI risks (employees sending sensitive data to unapproved LLM services), governance approaches like VPC deployments and API gateways, and how pipeline sprawl from complex feature engineering motivated Kumo's single-foundation-model approach with on-the-fly database queries. The discussion also covers how AI coding agents are changing engineering hiring practices, with interviews now focusing on understanding design choices rather than coding speed, and predictions that CIO/CISO tooling for AI egress visibility will become a must-have.