NVIDIA AI Podcast

Chris Wright, Red Hat CTO working on enterprise AI platform and governance, and Justin Boitano, NVIDIA VP overseeing AI infrastructure and blueprints. They explore the five-layer AI factory stack, agentic systems that plan and act, production-grade governance and security, hybrid cloud inference strategies, and starting small with validated designs and iterative evals.

Ben Sooter, Director of R&D at EPRI who works on innovation for reliable, affordable power, talks about how most AI energy is spent on inference rather than training. He describes micro data centers near underused substations to cut latency and boost resilience. He also covers distributed site clustering, grid integration strategies, and real-time low-latency applications.

Kuo Zhang, president of Alibaba.com and leader of its AI commerce efforts, explains how AI agents like Accio are transforming global B2B trade. He describes automating complex sourcing, compressing weeks of work into minutes, and enabling solo entrepreneurs and SMEs to access suppliers and design products. The conversation highlights agent orchestration, compliance across markets, and real-world use cases like large-scale event sourcing.

Marty Beard, CEO of Hayden AI, builds edge AI and camera systems for transit. Ahsan Baig, CTO of AC Transit, leads tech and innovation for the East Bay bus system. They discuss bus‑mounted cameras and NVIDIA‑powered edge AI that detect vehicles blocking lanes and stops. Topics include privacy‑by‑design, enforcement review workflows, early deployment challenges, rider accessibility gains, and expanding mobile AI for city asset management.

Dan Gural, head of technical partnerships at Voxel51 focused on data curation and workflows. Rohan Bhasin, senior solutions engineer at Fortellix specializing in sensor simulation and synthetic data. They discuss turning drive logs into high-fidelity simulations, neural reconstruction for faster scenario generation, and data-centric tools that prioritize edge cases and scale AV validation.

Jia Li, co-founder, president, and chief AI officer of LiveX AI, builds human-like AI agents and holographic experiences for retail, events, and travel. She discusses full-size 4K holograms that greet and guide people, real-time on-prem inference for crowded venues, Super Bowl and airport activations, and the technical challenges of latency, accuracy, and scaling multimodal models.

Nick Allardice, President and CEO of GiveDirectly, leads efforts to deliver digital cash to people in poverty and crisis. He discusses using AI, satellite imagery, and mobile money to forecast floods and preposition aid. Conversations cover rapid cash delivery within days, targeting damaged households with satellite data, and building trust and data partnerships for faster humanitarian response.

Jason Goldberg, known as 'Retail Geek', is the Chief Commerce Strategy Officer at Publicis Group and a seasoned expert in retail technology. He explores how AI is transforming retail by optimizing operations from supply chains to personalized shopping experiences. Goldberg discusses the rise of AI agents improving product discovery, the acceleration of AI deployments due to advancements in technology, and the potential disintermediation of retailers. He also highlights organizational hurdles in adopting AI and predicts a future where agents revolutionize retail transactions.

Brandon Hootman, Vice President of Data and AI at Caterpillar, discusses groundbreaking changes in heavy equipment through AI integration. He explains how digital twins and edge AI enhance safety and efficiency at job sites. The in-cab AI copilot exemplifies user-friendly technology aiding operators. Hootman also highlights Caterpillar's partnership with NVIDIA, utilizing real-time data to revolutionize supply chain management. As he looks to the future, he predicts a workforce augmented by AI and autonomous machines, reshaping industries.

Ian Buck, NVIDIA's VP of Hyperscale and High-Performance Computing, dives into the fascinating world of mixture-of-experts (MoE) architecture. He explains how MoE allows smarter AI models to operate with reduced compute costs by activating only necessary components. Buck highlights the importance of NVLink for maximizing performance and discusses the trade-offs between using MoE versus smaller models. Moreover, he envisions MoE's potential across various domains beyond just language, making a compelling case for its future impact.