Skip to content

From AGENTS.md to Enterprise Deployment

Practical AI

Hosted by Chris Benson and Daniel Whitenack

48m
Transcript

Original podcast

Where to listen

Summary

Written by a model from the transcript; not yet reviewed.

This episode of Practical AI features a discussion with a guest about bridging the gap between local AI agent experiments and highly regulated enterprise environments. The conversation covers how traditional platform engineering principles, such as those from Cloud Foundry and Tanzu, can be applied to containerize and deploy AI agents alongside applications. The speakers examine the challenges of air-gapped systems, compliance frameworks, model context protocols, and agent gateways, providing a blueprint for running agents securely and reliably at scale.

  • Enterprise environments often require strict compliance like PCI, SOX, or HIPAA and may even feature physical air gaps. 4:15
  • Using build packs simplifies container creation and deployment so developers do not need to manually configure infrastructure. 6:48
  • Model Context Protocol (MCP) gateways help control and scale access to approved tools and secure runtime services. 23:48
  • Enterprises should take baby steps by first securing access to an approved language model before scaling multi-agent systems. 33:09

Chapters

  1. 0:02
    Podcast Introduction and Host Greeting

    The hosts introduce the Practical AI Podcast, set the tone, and welcome listeners.

  2. 0:41
    Guest Introduction and Midwest AI Summit Preview

    Hosts introduce Nick and preview his upcoming talk at the Midwest AI Summit, including promotion details.

  3. 2:57
    Defining Enterprise AI Agent Environments

    Nick explains what constitutes an enterprise environment for AI agents, covering connectivity, compliance, and security.

  4. 4:59
    Enterprise Deployment Challenges and Solutions

    Discussion of regulatory hurdles, air-gapped setups, and how Tenzoo’s platform simplifies deployment and compliance.

  5. 7:00
    Memory, Identity, and Agent State Management

    Exploration of static versus dynamic elements in agents.md, memory services, and per-session identity handling.

  6. 8:40
    Security, Governance, and Monitoring of Agent Swarms

    Nick discusses guardrails, MCP gateways, and monitoring strategies to keep agent activity controlled and auditable.

  7. 37:00
    Organizational Adoption and Team Collaboration

    Strategies for integrating new AI teams with existing enterprise teams, breaking silos, and fostering cross-functional adoption.

  8. 44:40
    Future Outlook and Projections for Enterprise AI

    Nick shares his vision for the next decade of AI agents in enterprise settings, touching on scalability, regulation, and innovation.

  9. 46:01
    Automation Enhancing Productivity

    Personal reflections on how agents increase productivity and enable creators to build more.

  10. 46:33
    Voice, Robotics, and New Interaction Styles

    Speculation on voice integration, robotics, and future interaction paradigms with AI.

  11. 47:29
    Closing Remarks and Next Steps

    Hosts wrap up the episode, thanking the guest and promoting upcoming content and events.

Highlights

  • 4:12

    Enterprise Compliance Realitiesinsight

    Enterprises operate in highly regulated environments with strict compliance requirements like HIPAA and SOX, and some even have true air gaps requiring physical data center access.

  • 6:26

    Platform Engineering for Appsstrategy

    Platforms like Tanzu, based on Cloud Foundry, provide a standardized path to production by automatically handling container builds, security certs, load balancers, and app sandboxing.

  • 9:51

    Moving Agents Off Laptopsinsight

    Unlike local laptop development where stopping work halts an agent, enterprise deployment treats agents more like traditional applications that must run continuously in the cloud.

  • 12:45

    Co-locating Intelligence with Appsinsight

    To minimize latency issues at scale, enterprises prefer to co-locate large language models and agents as close as possible to the microservices and data they consume.

  • 17:43

    Deploying Agents via Build Packsstrategy

    Instead of writing raw container code, developers can push human-readable agent definitions using build packs that automatically provision them on the platform.

  • 21:49

    Automated Security Code Reviewsstory

    By connecting agents to repositories via webhooks, platforms can automatically trigger security and quality reviews on every single git commit.

  • 23:48

    Managing MCP Gatewaysstrategy

    MCP gateways abstract and regulate access to multiple MCP servers, allowing organizations to securely pass user credentials and track tool-use metrics.

Quotes

  • “There is no one AI kill switch.”
    Daniel Whitenack · 14:08
  • “You just have to exert that control across various layers of that stack.”
    Daniel Whitenack · 14:37
  • “Oh, oh, boy, it could be, it could be many things. But I think from a enterprise perspective,”
    Unnamed speaker · 45:07
  • “we're going to see like a radical change in like how, how, you know, development and, you know,”
    Unnamed speaker · 45:43
  • “I am like working harder than I ever have, because I have all these agents and they like, they need stuff from me.”
    Unnamed speaker · 46:01
  • “of, you know, what they can do for you. I think we'll see a lot with voice come, come into play,”
    Unnamed speaker · 46:33
  • Practical AI 1:24–1:28
  • Prediction Guard 15:03–15:38
  • Midwest AI Summit 30:05–31:27
  • Prediction Guard 48:32–48:36

About this episode

AI agents are moving beyond laptops and prototypes and into enterprise environments where security, compliance, scalability, and reliability matter. Nick Kuhn from VMware Tanzu Platform joins Daniel and Chris to discuss what changes (and what doesn’t) when deploying agents alongside traditional applications. They explore agent build packs, MCP gateways, shared memory, identity, sandboxing, and what enterprises can learn from years of platform engineering. Nick also shares practical advice for organizations adopting agents today and looks at what the future of AI could mean for enterprise teams and software development. Featuring: Nick Kuhn – LinkedIn https://www.linkedin.com/in/nick-kuhn-aa72a39/Daniel Whitenack – Website https://www.datadan.io/, GitHub https://github.com/dwhitena, X https://x.com/dwhitenaChris Benson – Website https://chrisbenson.com/, LinkedIn https://www.linkedin.com/in/chrisbenson, Bluesky https://bsky.app/profile/chrisbenson.bsky.social, GitHub https://github.com/chrisbenson, X https://x.com/chrisbensonSponsors: Midwest AI Summit: Join AI practitioners on October 15 in Indianapolis for practical sessions, hands-on discussions, and real-world AI solutions. Use code PracticalAI20 to save 20% on your registration. https://midwestaisummit.com/#ticketsPrediction Guard: A self-hosted AI control plane for running agents in high impact environments. predictionguard.com/practicalai https://predictionguard.com/practicalaiResources and Events: Register for upcoming webinars here https://practicalai.fm/webinars! Prior Webinars from our partner Prediction Guard https://www.youtube.com/playlist?list=PLRpRdFmBfdF6boDgAD3nNaLm8Yx4KN2VtMidwest AI Summit 2026 https://midwestaisummit.com/

Transcript

0:00
48:52

9,806 words · 4 speakers

Speaker names marked are inferred from speaking time and the show’s known hosts — they are not voice-verified.

Topics

  • agents
  • MCP
  • MCP servers
  • Tenzoo
  • agents.md
  • VMware
  • Enterprise environment
  • MCP gateway
  • hugging face
  • Midwest Summit

Page updated · transcript and summary are machine-generated