The Everything Feed - All Packet Pushers Pods TCG063: Constraint Drives Innovation with John Capobianco
Nov 26, 2025
Join John Capobianco, a technology practitioner and educator specializing in network automation and AI, as he dives into the vibrant AutoCon4 atmosphere. He shares insights on the hands-on RAG workshop, revealing attendees’ enlightening moments. The discussion contrasts RAG complexity with the streamlined Anthropic MCP. Explore how AI simplifies network automation tasks and the generational shifts in tech mindsets. John also touches on Google File Search's simplicity, the innovative TUN encoding, and the environmental considerations of large-scale AI models.
AI Snips
Chapters
Transcript
Episode notes
Workshop Moment That Turned Heads
- John ran a hands-on workshop teaching RAG then MCP and saw attendees light up when results matched real network data.
- One attendee asked for a router ID and the demo returned the correct value, which made the tech tangible for them.
Teach The Full Pipeline First
- Start tooling by walking through the full RAG pipeline to understand friction points before switching to newer abstractions.
- Use small, practical labs so teams can type, run code, and see immediate results they can apply to their work.
MCPs Collapse RAG Complexity
- MCPs drastically reduce friction compared with DIY RAG by removing vector-store, embedding, and orchestration complexity.
- Attendees quickly moved from code to natural-language interactions using MCP-enabled tools like Claude Desktop and Gemini CLI.