Scaling agentic AI: From infrastructure to enterprise outcomes – Kevin Cochrane
summary
Argues enterprises must move beyond experimentation to measurable AI outcomes by building full‑stack AI infrastructure using NVIDIA Dynamo and Nemotron 3 on globally distributed, composable cloud platforms (Vultr). Proposes API‑driven, prebuilt composable stacks that codify GPU, specialized storage and networking alongside CPU orchestration, enforce data residency and compliance, and enable one‑click scaling from 1 node (8 GPUs) to 512 nodes (~5,000 GPUs). Gives a hotel chain customer‑experience example and quantifies the operational need for standardized, reusable infrastructure templates and automation.
AI-written summary · about
chapters (8)
- 0:00Welcome
- 2:43Cloud-native to AI-native apps
- 6:54AI infrastructure super cycle
- 11:53Abstracting infra for devs
- 16:09AI-native hotel apps example
- 21:08Automating GPU clusters
- 24:20Global GPU deployment
- 26:47Conclusion
related talks
28mScaling agentic AI: From infrastructure to enterprise outcomes
12mDigital & Data/AI Strategy and Transformation
06mSponsored Keynote: Building Your AI Factory: Wh...
12mKeynote: AI Infrastructure
52mSimplifying the adoption of generative AI for enterprises
17mAI TechOps: Data and Infra for AI Foundation to Scale and Manage 1000s of Cohesive Agents
24mKeynote: Next-generation frameworks for Large-scale AI
53mHow to Build and Scale Ultra-Lean Gen AI Startups