Building Your First RAG Application by John Capobianco

A candid, fast-paced panel on what’s next in agentic AI and MCP—over the next 6 months and beyond. Leaders from across networking and AI – John Capobianco (Selector), Greg Freeman (Lumen), Du’An Lightfoot (AWS & Lab Every Day), Jeremy Shulman (MLB), William Collins (Itential), & Scott Robohn (Solutional) – discuss how to move from promising demos to productionized agents, what real trust and safety look like, why MCP’s careful evolution (incl. recent security work) matters, and where multimodal ops (think meeting-aware agents via Zoom RTMS) fit. Expect straight talk on determinism vs. non-determinism, scaling patterns, and why “agent on the box” isn’t as near as controller-driven, MCP-first workflows.

What You’ll Learn:
Near-Term Roadmap (Next 6 Months): Practical milestones for MCP adoption, security hardening, and taking internal agent tools customer-facing.
From Lab to Production: Patterns for deploying and scaling agents, handling guardrails, and building operator trust (including hybrid “deterministic wrapper + LLM” approaches).
Architecture Choices That Last: Why MCP-fronted controllers beat “AI on every router,” where A2A (agent-to-agent) fits, and how to think about standards vs. cloud-speed innovation.
Multimodal Ops in Practice: How voice/vision agents and meeting streams (e.g., Zoom RTMS) can observe, summarize, and even trigger actions—without humans living in the CLI.

If you wish to watch the entire day of AI for Network Leaders click here.

Watch the recording here

Explore the Selector platform

Try Selector now