← Home
Practical AI · May 28, 2026 · 48m

Rebooting Enterprise AI with MCP and Kubernetes

Chris Benson, Craig McLuckie, Daniel Whitenack
Show Notes

Craig McLuckie, CEO of Stacklok, discusses the infrastructure shift enabling AI agents to function as coworkers rather than chatbots. The conversation covers MCP (Model Context Protocol), Kubernetes orchestration, identity management, and the architectural patterns emerging for enterprise AI systems where fleets of agents operate autonomously behind the scenes.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Curious

•
Kubernetes Patterns for Agent Orchestration
Container orchestration patterns developed for traditional microservices (Kubernetes) are being adapted to manage fleets of AI agents, treating agents as scalable, stateless workloads.
ClaudeChatGPTGemini

Novel

•
MCP (Model Context Protocol) as Infrastructure Standard
MCP is emerging as a standardized protocol for how AI models and agents communicate with external tools, data sources, and other systems—similar to how HTTP standardized web communication.
ClaudeChatGPTGemini
•
ToolHive: Marketplace for Agent Capabilities
ToolHive is a centralized marketplace concept where enterprises can package, version, and distribute tools that AI agents can discover and use on demand.
ClaudeChatGPTGemini

Highlights

•
AI Agents as Coworkers, Not Chatbots
The architectural shift from AI systems that respond to queries to autonomous AI agents that operate continuously in the background, making decisions and taking actions with minimal human intervention.
ClaudeChatGPTGemini
•
Identity and Permission Management Becomes Critical
As AI agents gain autonomy to execute actions, organizations must implement permission models that control what each agent can do, similar to user access control in traditional systems.
ClaudeChatGPTGemini

Misc

✧MCP emerging as standard protocol for AI agent communication and tooling
✧Identity and permission management becoming critical as agents gain autonomy
✧Kubernetes patterns being adapted for agent fleet orchestration
✧ToolHive concept: centralized marketplace for agent capabilities
Was this useful?