Scaling PRs: Agentic Governance

Year: 2026 · ▶ Watch on YouTube

Karl Weinmeister (Director, Developer Relations) · Jeff Welpley (CTO & Co-founder / Google Developer Expert, Cloud & AI)

Switch language → zh

Segments (9)

  • 00:00:00 · Introductions — Karl Weinmeister, Jeff Welpley
    • Karl Weinmeister from Google Cloud introduces Jeff Welpley, CTO of GetHuman and a Google Developer Expert.
  • 00:00:27 · Topic Introduction: AI Agents and Multimodality — Karl Weinmeister
    • Karl sets the stage for a discussion on the intersection of AI agents, human communication technologies, and multimodal AI.
  • 01:01:00 · First Impressions of Next ‘26 — Jeff Welpley
    • Jeff shares his excitement about Gemini Enterprise and the agent platform, highlighting Google’s investment in making Google Cloud a hub for autonomous agents.
  • 01:48:00 · From Prototypes to Production — Karl Weinmeister
    • Karl discusses the industry’s shift from AI agent prototypes to dependable, production-level applications and the new SDLC challenges this creates.
  • 02:48:00 · Enterprise Challenges with Agentic Development — Jeff Welpley
    • Jeff outlines key challenges in enterprise AI adoption, including developer adoption rates and maintaining compliance and security standards with high-velocity, AI-generated code.
  • 04:25:00 · Governance Models for AI Development — Karl Weinmeister, Jeff Welpley
    • The speakers discuss governance strategies, such as senior staff oversight and using agentic systems for policy enforcement, with Jeff emphasizing ‘guardrails’ and reusable ‘skills’.
  • 06:02:00 · The Dual Meaning of ‘Skills’ — Karl Weinmeister
    • Karl clarifies that ‘skills’ refers to both upskilling developers and the newly announced ‘Google Agent Skills’ product for reusable agent functions.
  • 06:52:00 · Overcoming Developer Hesitation — Jeff Welpley, Karl Weinmeister
    • Jeff and Karl share anecdotes about how forcing themselves and their teams to use new AI tools for a trial period leads to a permanent shift in workflow and productivity.
  • 09:24:00 · Closing Remarks & Resources — Karl Weinmeister, Jeff Welpley
    • Karl points viewers to the Next ‘26 Codelabs, and Jeff gives a shout-out to the Angular community and a former colleague who spoke at the keynote.

Products Announced (1)

  • 06:15:00 · Google Agent Skills (Announced)
    • Provides reusable functions for AI agents · Helps enforce standards and best practices · Jump-starts development with pre-built capabilities
    • Announced at Google Cloud Next ‘26.

Customer Stories (1)

  • 00:00:18 · GetHuman — The company co-founded by speaker Jeff Welpley, which serves as the context for his experience with AI development.

Notable Quotes (4)

  • 01:10:00 — Jeff Welpley:

    I think Google’s put a lot of effort and obviously a lot of investment into making Google Cloud kind of the premier place to host agents, autonomous agents for your enterprise.

  • 01:49:00 — Karl Weinmeister:

    We’re in this new age with AI agents where it’s going beyond the prototypes.

  • 08:20:00 — Karl Weinmeister:

    I just told myself, I’m going to use Antigravity, nothing but it for one week and then, you know, kind of burn the boats behind me.

  • 09:08:00 — Jeff Welpley:

    I’ve done like a 180. Like I’ve changed my thinking dramatically and I cannot see myself ever going back.

Visual Signals

On-screen (4)

  • 00:00:00 · Google Cloud Next '26
    • Identifies the event and year on the background wall and microphone flags.
  • 00:31:50 · Karl Weinmeister Director, Developer Relations, Google Cloud
    • Identifies the host speaker and his role.
  • 01:06:30 · Jeff Welpley Google Developer Expert, Cloud & AI
    • Identifies the guest speaker and his affiliation as a GDE.
  • 10:21:50 · Google Cloud Next '26 logo
    • End slate for the video segment.

Stage (2)

  • 00:00:00 · The segment is a sit-down interview between two speakers at a desk, likely in a studio or booth at the Google Cloud Next conference.
  • 00:11:40 · The camera switches from a single shot of the host to a wide two-shot showing both speakers at the desk.

Key Topics

AI Agents · Enterprise AI · Autonomous Agents · Software Development Lifecycle (SDLC) · DevOps for AI · Multimodal AI · Gemini Enterprise · Google Agent Skills · Developer Productivity · Code Generation · AI Governance · Compliance and Security · CTO Perspectives · Developer Adoption · Google Developer Experts (GDE)

Takeaways

  • The enterprise AI landscape is maturing from experimental prototypes to business-critical, production-grade autonomous agents.
  • Google is heavily investing in making Google Cloud the primary platform for enterprise agents, highlighted by Gemini Enterprise and the new ‘Google Agent Skills’ product.
  • The high velocity of AI-generated code necessitates a re-evaluation of the entire SDLC, with a focus on new governance, compliance, and code review workflows.
  • Overcoming developer resistance to new AI coding tools is a key adoption hurdle; a ‘forced trial’ period can be an effective strategy to demonstrate productivity gains and create new habits.
  • Governance for agentic development is crucial, often managed through senior staff oversight or by building ‘guardrails’ and reusable, standardized ‘skills’ for agents to consume.
  • The concept of ‘skills’ is twofold: upskilling the workforce to use AI tools effectively and creating a library of pre-approved functions (like Google Agent Skills) for agents to use safely.
  • Even experienced developers are finding that fully embracing AI-powered development tools like the Gemini CLI can cause a permanent, positive shift in their productivity and workflow.