Your agents are starving! Airbyte’s Michel Tricot on the data ingestion crisis

About this episode

If your AI agents are starving for accurate context, or breaking production because they have way too much access, you are officially in the middle of the data ingestion crisis. This week, Airbyte CEO and co-founder Michel Tricot unpacks the massive data access and context sprawl challenges that are crippling AI workflows in production. He and Andrew dive into the critical need for strict data permissions, automated entity resolution, and why the transition to agentic workflows perfectly mirrors the paradigm-shifting migration from bare-metal servers to the cloud. Finally, Michel explains why obsessing over short-term AI ROI is a massive mistake, revealing how Airbyte measures true success with their internal automation project, Hydra. Get the guide: Your software factory needs a context layer Follow the show: • Subscribe to our Substack  • Follow us on LinkedIn • Subscribe to our YouTube Channel Follow the hosts: • Follow Andrew • Follow Ben • Follow Dan Follow today's guest: • Airbyte: Explore the open-source data integration platform powering AI context at airbyte.com • Agent Blueprint: Read Michel Tricot's newsletter on building agentic data infrastructure on Substack • Co