Moments after the opening keynote for VAST Forward, theCUBE’s Dave Vellante and Rebecca Knight set the stage from the company’s first-ever customer conference – a milestone moment for a 10-year-old company aiming to evolve from storage disruptor into something far more ambitious. With roughly 1,200 attendees and a partner lineup that reads like an AI infrastructure roll call – including Cisco, HPE, CoreWeave, Nebius, Lambda and Crusoe – Vellante and Knight frame VAST’s bet on a universal, software-based data platform built to serve AI at scale, across formats and across environments.

Vellante and Knight dig into what it really means to call VAST the “AI Operating System”: not a metaphor, but a claim about orchestrating data, security and IO as agentic systems move from demos to enterprise ROI. They also surface the harder questions emerging alongside the buildout – from zero-trust expectations and auditability to how humans will adapt when “agents per employee” becomes a real operating model. The result is a clear-eyed look at where AI infrastructure is heading, and why trust, transparency and platform architecture may determine who wins the next phase.

In this exclusive from VAST Forward in Salt Lake City, Jeff Denworth, co-founder of VAST Data, joins theCUBE’s Rebecca Knight and Dave Vellante to mark a decade of disruption – and to signal a far bigger ambition ahead. With more than $500 million in ARR and billions in deferred revenue commitments, Denworth describes a company riding an exponential curve, propelled by AI’s insatiable demand for flash, compute and globally distributed data infrastructure. What began as a storage rethink has evolved into what VAST now calls an “AI Operating System” – a software-defined platform spanning storage, database, compute orchestration and policy control. The goal: to underpin a $200 billion AI-driven infrastructure market by converging HPC roots with enterprise AI at scale.

The conversation dives into VAST’s architectural philosophy: first-principles engineering, disaggregated shared-everything design and a refusal to chase incrementalism. Denworth details new efforts around agentic memory, embedded fine-tuning engines and a policy framework that governs state, security and observability across distributed AI systems. As enterprises move from experimentation to deployment, VAST is positioning its platform as the foundation for AI factories: enabling continuous training, auditability and global data access without compromising performance. With deep co-engineering ties to major AI labs and cloud builders, the company is betting that infrastructure – not models – will define the next decade of intelligence.

In this segment from VAST Forward 2026, John Mao, vice president of business development and alliances at VAST Data, joins theCUBE’s Rebecca Knight and Dave Vellante to discuss VAST’s evolution into a comprehensive AI operating system. Mao explains how VAST is building a strategic alliance ecosystem that bridges the gap between hardware and applications to serve as the foundational layer for the modern AI lifecycle. A major highlight of the discussion is the native integration with CrowdStrike, which brings best-in-class cybersecurity directly into the VAST pipeline to secure data access and manage the emerging security risks associated with autonomous AI agents.

The conversation also explores VAST’s partnership with TwelveLabs, a frontier model company specializing in video and computer vision. Mao details how this collaboration enables industries beyond media and entertainment, such as finance and retail, to extract real-time insights from video assets for fraud prevention and customer analytics. Additionally, Mao shares the expansion of the Cosmos community, which has grown from a developer forum into a robust ecosystem of channel and technology partners. By providing open-source blueprints and fostering deep collaboration, VAST aims to simplify the complexity of AI infrastructure and accelerate enterprise-wide adoption.

In this interview from VAST Forward, Jonsi Stefansson, general manager of cloud at VAST Data, joins theCUBE’s Dave Vellante and Rebecca Knight to discuss how AI is dismantling traditional assumptions about cloud architecture. Stefansson highlights the inefficiencies of "lift and shift" models and explains how VAST’s Polaris acts as an intelligent abstraction layer and global namespace. He reveals how organizations can drastically reduce data fragmentation – moving from dozens of redundant copies to a streamlined, manageable footprint across edge, on-premises and cloud environments.

The conversation explores how VAST is solving modern supply chain challenges by intelligently matching data to available GPU compute power across different regions. Stefansson breaks down the technical advantages of the VAST AI Operating System, including its ability to vectorize unstructured data on the fly and consolidate the entire AI pipeline into a single platform. From eliminating egress-cost hurdles to delivering significant data reduction ratios, the discussion underscores how VAST is enabling enterprises to build more efficient, flexible and cost-effective AI factories.

In this interview from VAST Forward, Randy Hayes, vice president of VAST Data, joins Rob Linger, vice president of information advantage practice at Leidos, and Josh Salmanson, vice president of cyber defensive practice at Leidos, to talk with theCUBE’s Dave Vellante about the critical intersection of cybersecurity and AI ROI. The experts analyze a shifting threat landscape where adversaries now breach environments in under 30 seconds, far outpacing traditional defensive alerting systems. Salmanson explains why modern security has evolved into a complex data science problem, requiring the ability to process over 100 billion events per week to close the gap between detection and response.

The conversation explores how VAST Data and Leidos are collaborating to eliminate "tool bloat" by consolidating disparate security products into a unified AI operating system. Linger introduces the "Calibrated Trust Framework," a strategic model designed to govern autonomous AI agents and ensure auditable, responsible deployments within national security environments. By leveraging GPU-based architectures and VAST’s disaggregated shared-everything architecture (DASE), the panel highlights how organizations can transition from reactive defense to proactive hunting and achieve "mean time to know" at machine speed.

In this interview from VAST Forward in Salt Lake City, Phil Manez, vice president of go-to-market execution at VAST Data, and Ace Stryker, director of AI and ecosystem marketing at Solidigm, join theCUBE’s Dave Vellante and Rebecca Knight to unpack a supply shock most enterprises can’t ignore: a flash and memory crunch being accelerated by AI. Manez argues the demand curve isn’t behaving like a normal cycle – it’s being rewritten by inference at scale, where new applications and agentic workflows keep generating fresh data and chewing through capacity faster than suppliers can forecast.

Stryker explains that this isn’t a typical boom-and-bust cycle. AI inference is driving sustained demand – from RAG vectorization to larger context windows and KV cache requirements – and new fabs take years and billions of dollars to build. The conversation then shifts to solutions, including how the VAST Amplify program and its Disaggregated Shared-Everything Architecture (DASE) help customers get more out of existing flash through lower protection overhead and advanced data reduction, enabling them to extend scarce NAND capacity while continuing to scale AI initiatives.

In this exclusive interview, Danny McGinniss, vice president of product management for Cisco’s Compute Business Unit, and John Mao, vice president of business development and alliances at VAST Data, join theCUBE’s Dave Vellante and Rebecca Knight to unpack why “AI Factory” has become more than a catchy label – and why integration is now the product. McGinniss frames the partnership as a practical response to a hard reality: Enterprises want end-to-end, validated infrastructure that gets them to production faster, with security posture and observability designed in, not bolted on. Mao adds the field view from reference architectures – networking is the lifeblood, complexity is rising and most organizations can’t afford to treat AI infrastructure like a bespoke science project.

As AI shifts from IT efficiency to enterprise reinvention, the guests argue for a system mindset – “weakest link” thinking across compute, network, data and security – and for designs that let customers fail fast on use cases without failing on foundations. Vellante connects the dots to a new playbook. Get on the experience curve, build centers of excellence and treat GPU capacity like a shared service so utilization – and ROI – can compound. The takeaway is that production-grade AI will reward the organizations that simplify complexity early and invest with intent.

In this interview from Vast Forward 2026, Eric Bermender, manager of data infrastructure and platform at Pixar Animation Studios, joins Andy Pernsteiner, field chief technology officer at VAST Data, and Ace Stryker of Solidigm to talk with theCUBE’s Dave Vellante about the intersection of high-performance storage and cinematic innovation. Bermender explores Pixar’s technological evolution, from the early days of "Toy Story" to the complex demands of modern AI-driven rendering. He highlights how Pixar uses VAST Data’s disaggregated shared-everything architecture to eliminate the trade-offs between performance and capacity, allowing artists to iterate faster on volumetric characters. The conversation dives into the challenge of balancing deterministic creative control with the probabilistic nature of inference engines and how a unified flash layer simplifies the data pipeline.

The panel also breaks down the collaboration between VAST and Solidigm to deliver cost-efficient, lightning-fast storage through optimized QLC NAND. Stryker and Pernsteiner discuss how they are removing infrastructure bottlenecks to provide guaranteed latency for real-time artistic workflows. Bermender also shares how Pixar is integrating AI for preemptive diagnostics to avoid storage hangs, ensuring that hardware remains a transparent enabler for the next generation of stylized storytelling. From managing petabyte-scale render farms to navigating global supply constraints, the discussion illustrates how Pixar is leveraging the VAST AI Operating System to bridge the gap between legacy storage and deep learning requirements.

In this interview from VAST Forward 2026 in Salt Lake City, Chen Goldberg, executive vice president of product and engineering at CoreWeave, joins theCUBE’s Dave Vellante and Rebecca Knight to discuss the evolution from "Cloud 1.0" to the purpose-built AI cloud. Goldberg, a founding member of the Kubernetes project, explains why the traditional hyperscaler model requires a fundamental rethink to meet the extreme performance demands of modern AI workloads. She highlights how CoreWeave is prioritizing observability and transparency across the entire stack – from compute to storage – to ensure resiliency and efficiency for large-scale training and inference.

The conversation also explores how CoreWeave’s engineering-led partnership with VAST Data is enabling high-throughput solutions, such as distributed caching mechanisms that optimize data delivery to GPUs. Goldberg delves into the democratization of AI, noting how enterprises in healthcare and finance are moving beyond experimentation to deploy production-ready applications via managed APIs. From the acquisition of Weights & Biases to the critical balance of accuracy, price and latency in inference, she provides a roadmap for how organizations can leverage specialized infrastructure to accelerate innovation and overcome the technical debt of legacy cloud environments.