As part of NVIDIA’s OpenUSD Insiders livestream series, our SVP of Innovation, Susan Foley; our VP, Global Head of Technology, Peter Altamirano; and our VP, Computational Creativity & Innovation, Emrah Gonulkirmaz, dove into AI-driven marketing, content creation and the future of agentic workflows. In conversation with NVIDIA’s Jamie Allan, Director of AdTech & Digital Marketing Industries, and host Edmar Mendizabal, they explored practical use cases for digital twins and NVIDIA Omniverse—complete with hands-on tips, real-world client examples and advice for organizations eager to embrace these innovations.

If you missed it, you can watch the full session below or keep reading for the key takeaways.

Digital twins are the new foundation for creative scale.

Unlocking business value today means taking control of your creative assets and processes. Digital twins—the hyper-accurate virtual models of products, characters and spaces—are quickly becoming the bedrock of that approach. Allan set the stage for the session by explaining, “A big part of what we’ve been doing is [figuring out] how to evolve the content supply chain for marketing content and ads. A lot of that is founded in creating digital twins of products, whether it’s a car or a shampoo bottle. That’s where the power of OpenUSD comes into play.”

Digital twins are built using open standards like OpenUSD—an open-source framework and file format for describing, composing and interchanging 3D scenes and assets—and applications that are developed with platforms like Omniverse. They serve as the single source of truth for everything from product imagery to complex industrial simulations, allowing businesses to rapidly iterate, test changes virtually and deliver new products or updated assets in a fraction of the traditional time. As Altamirano said, “You can optimize layout, workflows, asset creation and test and simulate your processes far faster than in the real world—no matter if you’re updating a retail shelf, visualizing packaging or piloting new robotics workflows.”

That said, what usually holds organizations back isn’t a lack of understanding of the benefits of digital twins, but simply not knowing where to start. Foley advised, “Start small, scale up as you prove value, and we’ll help you migrate services to compute so you can build your own moat with owned intelligence.” Pro tip: Don’t let complexity slow you down. Today’s SDKs, libraries, templates and demo projects make it easier than ever to get started quickly with no need to build everything from scratch.

Modular, agentic workflows mean AI is now your creative partner.

The future of creative production isn’t simply generating more and more images and text with AI. It’s about orchestrating a system where specialized AI agents collaborate across the full pipeline. A standout example is our experimental AI-generated campaign for PUMA, where every stage—from initial script and storyboard to animation and editing—was orchestrated by AI agents using Monks.Flow, our professional managed service powered by AI. 

Thanks to NVIDIA NIM microservices and node-based orchestration enabled by Monks.Flow’s Pathways framework, AI agents can swap out models or creative roles as needed. Pathways uses self-learning AI to autonomously manage, optimize and adjust workflows in real time. For example, we can switch from one generative model for texturing to another for background imagery without disrupting the flow. 

Crucially, this entire process was anchored to a high-fidelity digital twin of the PUMA product, built using NVIDIA Omniverse libraries. “We started by importing a precise 3D model of the sneaker created in the OpenUSD format into NVIDIA Omniverse USD Composer,” explained Altamirano. This virtual product served as the foundational source of correctness for every subsequent creative step. 

The process didn’t stop there: synthetic data generated from the USD-based model was used to train and guide AI agents tasked with upholding the correct cinematic style throughout the film. That way, the team could control vital aspects required to meet brand guidelines, such as camera angles, lens choices and product accuracy.

As Emrah put it, “After bringing the product to the previsualization stage in Omniverse, we gained full control over the generation process.” Then, using Pathways, a network of specialized AI agents orchestrated animation, editing and scene composition, maintaining brand consistency throughout. 

How to get started and scale fast.

Whether you’re a marketer, a manufacturer looking to modernize, or a developer curious about AI-driven workflows, the new ecosystem emphasizes accessibility. “Get your first Omniverse setup done and you’ll see how reusable and scalable it really is,” said Altamirano. “Start with a prototype, prove the value, then expand. This approach works for retail, hospitality, even real estate.”

For organizations not sure where to begin, the advice is clear:

• Start small and show early results. Quick wins build stakeholder buy-in and reveal practical value.
• Invest in training not just for engineers, but for your whole team. NVIDIA Omniverse Blueprints, free Deep Learning Institute courses, and a vibrant developer community enable rapid learning and onboarding.
• Embrace open, modular platforms. This lets you change direction, upgrade AI models, and keep workflows on the cutting edge without locking yourself into monolithic systems.

Ultimately, modern creative innovation isn’t about one-off experiments; it’s about embedding intelligence, agility and modularity into the heart of your business. Digital twins anchor accuracy and scale. Agentic AI workflows make creativity collaborative and customizable. The path to scalable, AI-driven content creation has never been clearer. 

“We’ve built on our technical expertise to create an agile and efficient approach to Applied AI with a more nimble team of engineers and advisors.” says Michael Dobell, EVP of Innovation at Monks. “In today's landscape, a simple structure can put us one step ahead, allowing us to deliver with a speed and depth that legacy holding companies and networks struggle to match, setting a new standard for what’s possible in the industry.”

GTC attendees can see how agentic AI, using Monks.Flow, integrated with NVIDIA Omniverse and NVIDIA NIM™, creates a high-quality 30-second creative spot in a fraction of the time required by traditional production methods. By developing with Omniverse technology and leveraging NVIDIA Blueprints, Monks can orchestrate interoperability between creative teams and tools, allowing for scalable real-time development of hyper-realistic content. By harnessing the capabilities of NIM microservices on performant hardware, Monks was able to accelerate delivery of agentic workflows by over 2.8x. In addition, NIM microservices perfectly slotted into a cloud-agnostic strategy, utilizing the Kubernetes deployment strategy to easily run all of the backing models in all major cloud providers, for any client needs. Attendees can explore this innovation at the Monks booth in the GTC AI Pavilion, while audiences at home can now view a fully AI generated PUMA Film showcasing the incredible creative output.

This end-to-end workflow demonstrates how cutting-edge AI, coupled with Omniverse's capabilities, is transforming the future of content creation. By consolidating traditionally segmented areas—insights, strategy, creative, execution, media, performance and assets at scale—into a seamless, end-to-end agentic pipeline, the workflow results in work that is more adaptive, efficient and visually stunning—all at unprecedented speed.

“NVIDIA Omniverse and OpenUSD are enabling developers to build AI applications that assist artists and creative teams with efficient and scalable production techniques for digital marketing assets,” says Richard Kerris, Vice President of Media and Entertainment at NVIDIA. “Monks.Flow’s integration demonstrates how AI and digital twins can transform industries and set new benchmarks for innovation.”

At GTC this week, Lewis Smithingham, EVP of Strategic Industries at Monks, will present “Novel Use Cases for AI in Large Media Archives." Additionally, Tony Walasik, Senior Product Manager at Monks, will join Scott Connolly, Head of Media Technology at Verizon Innovation Labs, to discuss “AI on Location: Deploying Mobile Private Networks and Edge Compute for Next-Gen Content Pipelines.”

The Monks booth (#3263) offers attendees a more in-depth look at Monks.Flow, which leverages advanced AI models to transform businesses. By automating workflows and linking talent, software and microservices, Monks.Flow orchestrates marketing operations for the age of agentic AI, providing solutions for every stage of the creative lifecycle and offering a seamless journey from strategy to delivery. It begins with insights and strategy, where clients gain valuable data and develop targeted plans to guide their marketing efforts. This is followed by creation and adaptation, enabling the efficient production and customization of content to meet specific needs. Finally, Monks.Flow ensures delivery and performance, optimizing content to reach the right audience with maximum effectiveness. This platform has already delivered performance boosts for brands like Headspace, Hatch and Chime.

Our world is moving toward accelerated computing, with workloads being adapted into efficient AI workflows—including our Monks.Flow professional managed service, which connects exceptional talent with the greatest AI tools, APIs and models, all within easily orchestrated pipelines. Integrated with the latest innovations from NVIDIA, AWS and other technologies, we’ve built best-in-class solutions across each of our practices: Content, Data and Media. At GTC, we’ll demonstrate these solutions, and attendees can get a chance to see them in action.

Check us out at the NVIDIA XR Zone Pavilion – Monks.Flow and OpenUSD.

Stop by the NVIDIA XR Zone Pavilion to get a sneak peek at Monks.Flow, an AI-centric professional managed service that streamlines how humans and machines interoperate at every step of the marketing journey, from building business intelligence to delivering hyper-personalized creative content. Monks.Flow leverages NVIDIA CloudXR and the NVIDIA Omniverse OpenUSD development platform to help brands virtually explore different customizable product designs and unlock scale and hyper-personalization across any customer journey.

Stop by the AWS booth (#708) for more on Monks.Flow, implementations and its use cases.

In addition to the NVIDIA XR Zone Pavilion, you’ll also find our featured space at the AWS booth, #708. There, we’ll have a dedicated kiosk where you can meet the Monks and learn all about the AI solutions built on AWS and NVIDIA technology that we’re delivering to CMOs.

We have designed a state-of-the-art workflow that helps marketers make better decisions by allowing them to chat directly with their customer personas, which are enriched with a comprehensive knowledge base and multiple large language models. This workflow delivers a tailored, secure, and scalable AI language framework for clients. Providing an understanding of language, context, emotion, and tone, these demos showcase the types of dynamic, low-latency interactions made possible by AI.

Wormhole, our out-of-this-world animatronic alien, will be stationed at our kiosk, ready to chat with passersby. Connected to the cloud and powered by AWS AI tools, including Bedrock, Amazon Sagemaker and Amazon Polly, Wormhole demonstrates lifelike movement and speech and serves as a prototype for the kinds of human-to-robot creations that are made possible by advances in AI—with a touch of humor.

After chatting up Wormhole, stick around for a showcase of Novo, a lifelike virtual human. The ultra-realistic avatar integrates NVIDIA NeMo and NVIDIA Audio2Face, both part of the NVIDIA ACE suite of technologies, to deliver an immersive and engaging conversational experience.

Don’t miss informative talks throughout GTC.

How Media.Monks Builds Generative AI Experiences (March 20, 10am PDT)

If you want to learn more about Wormhole and how it all came together, don’t miss this talk led by Iran Reyes (VP, Global Head of Engineering, Media.Monks), Peter Altamirano (VP, Global Head of Technology, Media.Monks), and Gregoire Rouyer (Global Partner Strategy Lead, AWS). Taking viewers behind the scenes into the creation of Wormhole, the trio will discuss relevant AWS AI tools, including Bedrock, Amazon Sagemaker and Amazon Polly. Learn how it was built and pose your questions to the team!

Revolutionizing Fan Engagement: Unleashing the Power of AI in Software-Defined Production (March 21, 10am PDT)

Presented by Lewis Smithingham, SVP of Innovation and Creative Solution at Media.Monks, this talk will discuss how software-defined broadcast and Monks.Flow for Media and Entertainment are transforming broadcast and unlocking cost reduction, content scalability, talent acquisition, sustainability and more.

Our advanced live production pipeline is powered by NVIDIA hardware, including NVIDIA RTX 6000 Ada Generation GPUs and NVIDIA L40S GPUs, to support 8K 60fps pipelines for live broadcast. Media.Monks’ latest iteration of broadcast and content delivery tools integrates NVIDIA Holoscan for Media to serve hyper-personalized content to fans across new media formats. In order to build this content at speed and scale, Media.Monks utilizes the NVIDIA DeepStream SDK, part of the NVIDIA Metropolis platform for developing vision AI solutions, to identify in-frame images within specific broadcast segments. It also uses the NVIDIA NeVA model, hosted with the NVIDIA Triton Inference Server, which etches relevant metadata information to accurately deliver the segment to interest-based audience groups. Retrieval augmented generation (RAG) is used to aggregate and segment audience data sets, and, in turn, serve hyper-personalized content to viewers with incredible precision. Finally, the software-defined production pipeline brings together a series of technologies from AWS, including EC2 (instance types G4dn and G5), CloudFront, Route 53, CloudWatch, and CloudTrail.

Watch the presentation to learn how it all comes together to help brands and broadcast rights-holders reach incredibly specific and granular audiences—for example, a basketball viewer in Utah who likes rap music, collecting sneakers, and gaming.

AI Orchestration for the CMO with Monks.Flow

Take a look at Monks.Flow, our professional managed service for marketers that trades manual workloads for AI workflows. The platform-agnostic orchestration tool is designed to work across an organization’s existing tech stack and is built around the unique needs of the enterprise. Monks.Flow and our consulting practice help CMOs reduce costs, select tools and partners, develop model and owned data strategies, integrate teams, build business intelligence and maximize marketing impact.

Our latest Monks.Flow release is Persona.Flow, which provides deep consumer intelligence by simulating detailed, interactive consumer personas that are engaged in a question-and-answer format. Facilitated by a RAG framework that orchestrates brand and factual consumer data, Persona.Flow provides brands with the insights they need to fuel content generation and creativity at scale.

While we’ll have videos at the kiosk inside AWS booth #708 demonstrating the abilities of Monks.Flow, VP, Emerging Tech and R&D Sander van der Vegte and EVP, Innovation Michael Dobell will also be presenting on the solution at the nearby AWS In-Booth Theatre, located at the AWS booth. Stop by on March 20 at 1:50pm PDT for the formal presentation, or stop by at our kiosk any time to learn more. Viewers will learn firsthand how AI-powered workflows are enabling marketing-led transformation across the enterprise.

We can’t wait to see you!

As we approach NVIDIA GTC, we invite everyone to join us in exploring the incredible potential of AI and machine learning. From advanced workflows to immersive conversational experiences, we’re excited to provide a glimpse into how these technologies are transforming the industry now, not just in the future. We hope to see you there!