APIAI26 D2 WGB2 1530 B2 4 PGM 01

AI can make API prototypes look production-ready in minutes—but regulated enterprises still need governance, auditability, and approval gates before anything ships. This talk shows how platform engineering teams can combine Kong, APIOps, CI/CD validation, AI gateway patterns, MCP, and PR-based policy checks to reduce manual "human middleware" while keeping API compliance intact. KEY TAKEAWAYS • AI can generate demo-ready APIs quickly, but production deployment in regulated environments still requires auditable controls and human approval. • Platformizing API governance with Kong lets teams inherit reusable controls like authentication and SSO instead of revalidating every baseline policy manually. • AI coding assistants become more effective when connected to platform documentation and skills through an AI gateway and MCP server. • PR-based policy agents and deterministic APIOps checks create a feedback loop where AI can fix compliance failures before deployment. • Removing human middleware shifts compliance work from tickets, emails, and meetings to reusable policies that teams can apply consistently. TIMESTAMPS 00:00 - Intro 00:12 - AI-built APIs and risk 01:12 - Regulated enterprise context 02:19 - Policies, processes, audits 03:15 - Manual compliance bottlenecks 03:55 - Platformizing governance with Kong 05:28 - API lifecycle complexity 06:10 - Self-service API deployment 07:35 - AI skills in gateway 08:41 - Nondeterminism and governance 09:39 - Policy-based PR agent 10:35 - Deterministic APIOps checks 11:07 - Iterative AI remediation 12:21 - Faster regulated delivery 13:16 - Human approvals remain 13:41 - Compliance-owned policies 14:41 - Removing human middleware 15:36 - Q&A on scoped policies LINKS Explore how Kong helps scale and govern agentic infrastructure: https://konghq.com/solutions/agentic-... #APIGovernance #AgenticAI #APIOps #Kong #PlatformEngineering