The Everything Feed - All Packet Pushers Pods NAN102: Practical Applications for AI in Network Automation
Oct 1, 2025
In this engaging discussion, Ryan Booth, an expert in network automation with experience at Juniper and Abstra, shares his journey from network engineering to AI. He highlights the importance of community and persistence in success and his exploration of AI tools. Ryan delves into practical AI applications for network engineers, discussing retrieval-augmented generation and the importance of domain knowledge over just model building. He offers valuable insights on automating workflows and bridging gaps between network and software teams.
AI Snips
Chapters
Transcript
Episode notes
Start With Why Before Using AI
- Keep asking 'why' before adding AI to a workflow and validate the user value it creates.
- Resist adding chat or LLMs where simple code or process changes suffice.
AI As An Assistant, Not A Replacement
- Ryan views AI as an assistant that elevates human skill rather than a straight replacement.
- Assistive AI can simplify complex operational tasks and extend engineers' capabilities.
Tools And MCP Enable Agentic LLM Workflows
- Tool-enabled LLMs and the MCP standard let models call external APIs and orchestrate workflows.
- This opens richer agentic applications that connect LLMs to real systems and data.