CTO Circle: AI-Native Organizations Need Workflows, Not Just Tools

Snowflake hosted the inaugural CTO Circle during Snowflake Summit 2026 in San Francisco, convening more than 350 CTOs from financial services, telecommunications, retail, and technology. The event was created to provide engineering leaders a trusted forum to exchange lessons on AI-native transformation, moving beyond competitive isolation that typically keeps such conversations inside individual companies.

The discussion converged on three defining themes of AI-native engineering organizations. First, on AI in production, Vivek Raghunathan, SVP of Engineering at Snowflake, argued that treating developers as customers—rather than as a productivity or culture problem—transforms how organizations adopt AI. Snowflake interviewed developers to map friction points across the software development lifecycle, established baseline metrics, and ran experiments to measure the impact of changes. The approach increased the company's internal developer Net Promoter Score by more than 30 points in 18 months, achieving a 4:1 ratio of satisfied to dissatisfied developers. Raghunathan described three stages of AI adoption: Adoption (learning to use AI tools daily), Mastery (discovering repeatable workflows that produce better outcomes), and Optimization (making those workflows organizational knowledge). At Snowflake, early adopters experimented with prompting techniques and debugging approaches; the organization then documented successful patterns and distributed them across engineering, so every developer had access to the same tools but engineers using proven workflows consistently outperformed those at early adoption levels. Jon McNeill, author of The Algorithm, encouraged leaders to reverse the typical approach: instead of beginning with technology and searching for applications, successful organizations identify key business constraints and redesign engineering systems around solving them. Success, McNeill suggested, will be measured not by code generation or token consumption but by who builds the simplest, fastest engineering system and eliminates the most unnecessary steps between idea and production.

Second, on balancing velocity and risk, Jeremy Burton, General Manager of the Observability Business Unit at Snowflake, argued that AI has entered a new phase where early experimentation is giving way to production deployment and measurable business value. AI agents generate more telemetry and interact with more systems, changing observability from infrastructure monitoring to providing the context AI systems need to operate reliably. Burton challenged the assumption that better models simply need better data; in reality, AI is only as effective as the context it can access, which extends beyond raw telemetry to include data semantics, relationships through ontologies and knowledge graphs, and business context. Aditya Gaur, Engineering Manager at Netflix, demonstrated this principle through Netflix's work on automated root cause analysis, which succeeded not primarily through AI but through years of prior investment in connecting fragmented telemetry, modeling operational relationships through an ontology and knowledge graph, and creating a shared context layer. Caitlin Colgrove, CTO at Hex, urged leaders to commit fully to AI-native structures, what she called "burning the boats"—restructuring how teams build products and make decisions rather than creating isolated AI teams. Hex initially created a dedicated AI product team, which became a bottleneck; disbanding it and distributing AI responsibility across every product team allowed the company to deliver products rapidly with embedded AI. Chris Kozlowski, Managing Director at Barclays, emphasized that speed creates value only when matched with governance and trust, particularly in regulated environments. Corey Burke, SVP Engineering at Dialpad, and Arun Rajamanickam, VP of Engineering at project44, described how AI shortens the path from identifying a customer problem to validating a solution, but only if teams have infrastructure, shared context, and operational discipline to iterate confidently.

Third, on designing engineering teams for AI, Qi Jin, EVP at Cerebras Systems, suggested that organizations optimized for previous software development generations often become the biggest barrier to AI adoption, as established processes and boundaries were designed to scale proven ways of working rather than to continuously adapt. The underlying theme across all three areas was that sustainable competitive advantage comes from institutionalizing how AI is used—through workflows, data architecture, and team structures—rather than from deploying new tools alone.