NVIDIA AI Podcast — Ep. 293
Building AI Factories: How Red Hat and NVIDIA Turn Enterprise Data Into Intelligence
Release Date: March 12, 2026
Host: Noah Kravitz
Guests: Chris Wright (Red Hat CTO & SVP, Global Engineering) and Justin Boitano (NVIDIA VP & GM, Enterprise Computing)

Episode Overview

This episode delves into the concept of "AI factories"—purpose-built environments that transform enterprise data into actionable intelligence. Guest experts from Red Hat and NVIDIA discuss why enterprises are rapidly adopting these AI factories, what foundational components are needed, how to deploy and scale them with confidence, and what the future of enterprise AI looks like as agentic systems become central to business operations.

Key Discussion Points and Insights

1. Defining the AI Factory (00:50-03:25)

• Justin Boitano outlines the AI factory as the next industrial revolution for enterprises:

  “Building digital intelligence to power the productivity of organizations is going to be as critical in this decade as energy in running our companies. This is the next industrial revolution.” (01:24)

• The “five-layer cake” model of an AI factory:
  1. Data Centers & Hardware: Power and chips, now at rack scale.
  2. Software Infrastructure: Orchestration and management.
  3. AI Models: Running intelligence.
  4. Applications & Agents: User-facing outputs.
  5. Use-Case Driven Outcomes: Business-oriented solutions.

2. Bridging New and Legacy Enterprise Worlds (03:00-08:32)

• Chris Wright discusses blending new AI capabilities with existing enterprise systems:

  "A lot of what we're doing is taking these building blocks… and making them accessible to the enterprise together... building the right guardrails and security considerations... so our customers can feel confident about bringing this into their enterprise." (04:10)

• Emphasizes need for robust governance, audit trails, and access controls—often neglected in rapid AI tool adoption.
• Quote:

  “You need to bring AI capabilities not just in the net new, but also in the existing content that runs the enterprise. And to me that's exactly what the AI factory does. It helps bridge these two worlds.” (06:54)

• Lack of standardization leads to fragmented shadow IT and higher project failure rates; the AI factory approach builds consistency and best practices.

3. The Rise of Agents and Evolving Use Cases (08:32-11:30)

• Justin Boitano observes accelerating market interest and a shift from simple chatbots to sophisticated, autonomous agentic systems:

  “We can feel these agents doing so much more work for our developers and running longer, more complex software tasks... This moment of clause came out and clause basically take[s] it… to a new frontier of full autonomy.” (09:10)

• Significant cost savings with hybrid model architectures—NVIDIA sees up to 30x cost reduction in new blueprints by combining open models (on-prem) for most workloads with “frontier models” only for planning stages.

4. Security, Governance, and Scaling with Confidence (11:30-15:10)

• Importance of robust security, separation of development vs. production, and role-based access control:

  “The worst thing that enterprises can do is overanalyze this though... You've got to believe that AI is this new frontier and the companies able to harness it... are going to have a massive competitive advantage.” (13:33, Boitano)

• Production environments (inference) vs. training: inference is the real-world, scalable deployment of intelligence, and requires robust monitoring, cost optimization, and compliance.
• Hybrid cloud and edge deployment flexibility are enabled by shared, consistent AI factory footprints.

5. Building and Scaling the Factory: Infrastructure and Platform (18:02-24:28)

• Sensible stack sequencing: Starts with hardware and power decisions (e.g., air vs. liquid cooling), then software orchestration, model delivery, and application layers.
• NVIDIA Approach:

  “We provide reference blueprints, which are examples of proven use cases that even we run on our AI factories at NVIDIA... Enterprise search... then from there you can start to expand into your own developed use cases.” (19:36)

• Red Hat’s Role:

  “The stack starts with hardware, hardware enablement and... the distributed nature of rackscale architecture... Then Kubernetes for orchestration, and above that, Red Hat AI Enterprise layers with NVIDIA’s optimized models... We sometimes call this the metal-to-agent stack." (21:34, Wright)

• The tension: Early wins matter—“perfection is the enemy of good enough.” Pick use cases that deliver real business value and build iteratively.

6. The First 90 Days: Practical Steps for AI Factory Teams (24:28-28:58)

• Start with validated designs and reference blueprints for deployment.
• User Acceptance Tests (UAT): Early feedback loops with real users quantify productivity gains.

  “You can really quickly get to, from time savings across a user group to productivity gains... If you can get a 2x productivity gain across a big population of users, then, you know, you're onto something really big." (26:12, Boitano)

• Iterative, hypothesis-driven development:

  “Pick something that's real so it's not so artificial... Focusing on those things that are real, but again not making it too big... Right sizing and the iterative process of learning as you go is how you start building the thing that ultimately is quite big.” (28:00, Wright)

7. Guardrails for the Road Ahead (28:58-31:17)

• Embed security teams early, audit user and agent permissions.
• Treat agents as “digital employees” or “contractors,” with least-privilege access and staged onboarding.
• Evolving permissions and monitoring as agents become more autonomous.
• Quote:

  “You'll find AI is really good at doing discovery in business systems that it has access to... So, having the security teams understand, are we allowing too broad of access to what we want to keep confidential?" (30:00, Boitano)

8. Vision: The Future of AI Factories (31:17-36:53)

• Chris Wright:

  “Where agents take on critical tasks in the business... it's not that we're going to go through and augment each of today's processes... It's really redefining how we work together completely end to end.” (34:23)

• Predicts a shift in which AI factories underpin enterprise operations—autonomous, agent-driven processes become the core.
• Justin Boitano:

  "You're going to have different agents working for you that you give these more structured, long running tasks to. They go off and think and do the work and then they come back to check in in a period of time." (35:26)

• Forecast: Productivity for software engineers—and soon for knowledge workers across industries—will increase 2-3x, transforming work organization-wide.

Notable Quotes & Memorable Moments

• “Companies are always asking us, how do we build these factories that basically take data in and then produce the intelligence that helps them run their businesses more efficiently.” (01:35, Boitano)
• “A lot of what we're doing is… building the right guardrails… so our customers can… bring this into their enterprise as they're trying to… go from a traditional company to really an AI native company…” (04:07, Wright)
• "Perfection is the enemy of good enough." (23:23, Wright)
• "If you can get a 2x productivity gain across a big population of users, then you know you're onto something really big." (26:15, Boitano)
• "It's really redefining how we work together, completely end to end." (34:25, Wright)
• "Every company across every industry and every job function will really be transformed with the use of an AI factory." (36:46, Boitano)

Key Timestamps

• Defining AI Factories & Five-Layer Cake: (00:50–03:25)
• Integrating AI with Enterprise IT: (03:00–08:32)
• Trends in Agentic AI: (08:32–11:30)
• Security & Governance in AI Factories: (11:30–15:10)
• Stack/Platform Details: (18:02–24:28)
• First 90 Days, Early Wins: (24:28–28:58)
• Designing Guardrails: (28:58–31:17)
• Vision for the Future of AI Work: (31:17–36:53)

Resources & Further Learning

• Learn more about Red Hat AI Factory with NVIDIA:
  “That's sort of an easy thing to search and you'll find information from redhat.com… together with Nvidia on the Nvidia website...” (37:11, Wright)
  – Red Hat AI Factory with NVIDIA
  – NVIDIA AI Podcast

Summary Takeaway

AI factories are rapidly becoming the backbone of enterprise intelligence, closing the gap between legacy IT and state-of-the-art AI. With structured frameworks, trusted partnerships (Red Hat & NVIDIA), and focus on security, governance, and scalability, companies can transition to truly “AI native” operations—unlocking unprecedented productivity gains and transforming the very nature of work. The future is not decades away; it’s unfolding quarter by quarter—driven by agents, validated frameworks, and thoughtful, iterative adoption.