Panel: Agentic AI Infrastructure for Telecom
summary
Demo shows agentic lifecycle automation for CNF deployment at MWC and a security use case where cooperating agents ingest logs, generate rules and adapt on the fly; pilots run in two countries and a 60 million users pilot for live‑stream QoS control is live. Telecom stacks integrate MicroVM, Firecracker, StratoVirt, libkrun for strong isolation at edge and cloud, require sub‑10 ms POP latency targets, proxied MCP/A2A for carrier grade comms, and harden observability, context memory and cost controls before production.
AI-written summary · about
chapters (9)
- 0:00Agentic AI in telecom
- 1:04Autonomous lifecycle management demo
- 3:13Panelist use cases
- 13:09Deploying agentic AI challenges
- 16:44AI adoption hurdles
- 21:56LLMs versus simpler AI
- 24:04AI deployment standards
- 32:52Telecom agent infrastructure
- 36:18Closing remarks & Q&A
related talks
54mKeynote
32mIntelligent cloud networking: Create simplicity, remove complexity (PRT214)
31mAn Open Source Telco Cloud Stack to Accelerate the Adoption of Cloud Native Operator Model
13mKeynote: Agentic AI Is Running Your Infrastructure
17mSponsored Keynote: AI-Driven Deployment and Monitoring Across Federated Cloud-Edge Sites
56mOpening Keynote
44mMCP Security: Keep Your AI Agents from Spilling the Tea
42mAI for security - friend or foe?