Episode summary
Almost every enterprise on Earth has an AI project running — but many will never make it into production. Not because the models don’t work, but because the data underneath them isn't ready for AI. In this episode of The Data Wire, Everpure’s Shawn Rosemarin sits down with NVIDIA SVP of Networking Kevin Deierling to unpack why data intelligence at the source has become the real prerequisite for enterprise AI, and how AI‑ready data and AI factories turn that intelligence into durable competitive advantage. (Watch the livestream of their session at Accelerate 2026 at…
Chapters
- 00:00 — – The uncomfortable truth: why most enterprise AI projects never reach production
- 01:07 — – Meet NVIDIA’s Kevin Deierling, SVP of Networking
- 03:29 — – “The compute works”: GPUs aren’t the bottleneck, unrefined data is
- 04:10 — – Defining AI‑ready data and data intelligence at the source
- 07:07 — – From factories to AI factories: data as the raw material of business value
- 07:39 — – Agents, reasoning loops, and AI “eating its own tail” at GPU time scales
- 09:48 — – How bad, stale, or conflicting data quietly kills AI projects
- 11:27 — – Why NVIDIA partnered with Everpure on an AI‑native, data‑intelligent platform
- 12:38 — – Beyond ChatGPT: agentic AI and the data intelligence bar it sets
- 14:49 — – “You absolutely cannot wait”: compounding productivity gains from AI
- 17:13 — – Refined, governed AI data as a durable advantage competitors can’t buy
- 20:15 — – Teaser: what Everpure + NVIDIA are unveiling at Accelerate
- 20:51 — – “AI is a team sport”: compute, network, and data as one system
Leave a Reply