981: How Data Engineers Are “10x’ing” Themselves With Agents, feat. Matt Glickman

In this episode of Super Data Science, Jon Krohn interviews Matt Glickman, a veteran data leader with nearly 25 years at Goldman Sachs and a key early adopter of Snowflake, about his new venture, Genesis Computing. Glickman describes February 2026 as a pivotal 'event horizon' moment when AI models achieved the capability to handle complex, multi-step data engineering workflows that previously required deep human expertise. He explains how Genesis Computing is building an agentic platform that automates data engineering by creating a 'living context graph'—a dynamic, AI-mapped representation of an organization’s data relationships, documentation, and business logic. Unlike traditional co-pilots, Genesis agents work autonomously, only escalating to humans when confidence is low, and they continuously learn and memorialize decisions to prevent knowledge loss. The platform is deployed on-premises or in private clouds, ensuring enterprise security and data ownership. Glickman emphasizes that AI is not replacing data engineers but 10x-ing their impact, eliminating the need for junior hires while creating new roles for 'AI orchestrators.' He also discusses the shift from 'AI use case' thinking to asking 'why can't an agent do this?' and outlines a four-phase adoption model from assessment to scaled autonomy. The episode concludes with reflections on AI’s potential to discover new knowledge, citing experiments like training models on pre-Einstein data to recreate relativity.