Approach to AI at a large scale
summary
Worldwide Technologies describes a repeatable end to end approach: 14 years of experience, labs using F5 for security, multi‑tenancy and high performance transport, and a testing/certification pipeline that surfaces deployment gaps. Infrastructure work moves from design to governance, automation and factory deployment, with GPUs and vector DBs scaling a RAG workload from 1 to 15 GB per query and network links rising from 100G to 800 gig and beyond toward 1.6T. BlueField NIC offloads and BIG‑IP integration drive the measurable transport and security improvements.
AI-written summary · about
related talks
06mManaging and Securing AI Output
09mAI for NI
34mRAG AI Deployments
17mAI TechOps: Data and Infra for AI Foundation to Scale and Manage 1000s of Cohesive Agents
28mScaling agentic AI: From infrastructure to enterprise outcomes
11mAccelerating and Streamlining LLM AI Deployments- 28mScaling agentic AI: From infrastructure to enterprise outcomes – Kevin Cochrane
12mKeynote: AI Infrastructure