Bret and Nirmal explore AI agents in Kubernetes with Eitan Yarmush, Senior Architect at Solo.io. Eitan explains how AI agents work through three simple components (system prompts, LLMs, and tools), and demonstrates the kagent project, which provides a Kubernetes-native way to deploy and manage AI workflows.
🙌 The Agentic DevOps Guild has launched! It's a training + community + mentorship program for engineers wanting to learn the latest CI/CD automation and dive into Agentic DevOps. Meetups are happening now, with new course videos dropping every few weeks. Join the Guild and become your team's leader in AI for infrastructure automation https://www.bretfisher.com/theguild 🍾
Also in this episode, we cover the Model Context Protocol (MCP), which standardizes how agents communicate with external tools, along with practical use cases like incident response, debugging workflows, and CI/CD integration. Eitan shows how to create an agent through a web interface.
We also address the challenges including security concerns (MCP lacks built-in security standards), cost considerations, and technical limitations like context window constraints.
(04:36) - Hype vs Reality
(06:28) - Incident Response and Debugging
(09:46) - Defining Agents and MCP
(15:33) - Where Do You Run Agents?
(18:37) - What Problem does kagent Solve?
(29:05) - Authentication and Tool Limitations
You can also support this podcast by subscribing to my YouTube channel and my weekly newsletter at bret.news!
Podden och tillhörande omslagsbild på den här sidan tillhör
Bret Fisher. Innehållet i podden är skapat av Bret Fisher och inte av,
eller tillsammans med, Poddtoppen.
