Presented by Nutanix
Organizations across diverse industries are grappling with the challenge of transitioning artificial intelligence (AI) initiatives from preliminary stages—like pilot programs, proofs of concept, and cloud-based experiments—to full-scale deployment within their operational environments. VentureBeat engaged in a discussion with Tarkan Maner, President and Chief Commercial Officer at Nutanix, and Thomas Cornely, EVP of Product Management, to explore the essential requirements for this crucial leap and the strategies needed for successful execution.
“AI is fundamentally reshaping every aspect of our operations, not just within the technology sector but across all vertical industries,” stated Maner. “This transformation spans regulated sectors such as banking, healthcare, and government, alongside less regulated fields like manufacturing and retail. As a comprehensive platform company, we view this evolution as a significant opportunity to enhance how we serve our clientele moving forward.”
However, Cornely highlighted that a practical chasm persists between the phases of experimentation and production. “Conducting an experiment or developing a prototype is distinct from deploying that prototype to serve, for instance, 10,000 employees,” he explained. “The focus has shifted from merely training models or building chatbots to developing sophisticated agents, which places exponentially growing demands on AI infrastructure.”
Agentic AI introduces a new layer of enterprise complexity
The emergence of agentic AI is a pivotal factor in this transition, introducing sophisticated multi-step workflows that span various applications and data sources. Coupled with a degree of autonomy, these systems necessitate novel operational capabilities and stringent oversight.
Enterprises must now manage the simultaneous operation of multiple agents, accommodate unpredictable, real-time workloads, and ensure coordinated access to infrastructure resources across disparate teams. “Tools like OpenClaw are significantly simplifying the process for anyone to create and deploy agents,” Cornely noted. “Crucially, these agents should operate on-premises, leveraging your sensitive data. Robust security constructs are essential to safeguard the enterprise from any potential actions by these autonomous agents.”
As these systems gain greater autonomy, the focus extends beyond their operational mechanics to their intricate interactions with enterprise data, existing systems, and human teams.
AI is augmenting human work, not replacing it
Maner emphasized that agentic AI serves as a powerful enhancer of human capabilities rather than a direct replacement. The strategic objective for enterprises is to achieve an optimal equilibrium between human oversight, AI-driven automation, and agent-assisted workflows, rather than aiming for complete human displacement. “We envision a harmonious coexistence between AI, agentic tools, robotics, and human talent,” Maner asserted. “This synergy can be expertly optimized to yield superior business outcomes for enterprises, governments, and public sector organizations, provided the right vendors deliver the requisite tools and services.”
How enterprises are getting started with AI at scale
In practical terms, the journey from initial experimentation to tangible, real-world deployment reveals the most significant hurdles. Despite considerable momentum, many organizations are still navigating the complexities of scaling AI beyond its nascent applications. During this process, practical limitations invariably surface. While many initiatives commence in the cloud due to the immediate availability of resources and services, critical factors such as data management, governance, control, and cost quickly become paramount concerns.
The cloud serves as an effective environment for experimentation, with the ultimate objective of migrating applications back on-premises as they mature into production, utilizing platforms that effectively address security and cost considerations. Prominent use cases currently gaining traction include enhancing document search and knowledge retrieval, bolstering security through predictive threat detection, streamlining software development and coding workflows, and optimizing customer support operations. In the security domain, financial institutions and other clients in Europe and the U.S. are actively deploying AI-powered tools, including facial recognition and advanced threat prediction mechanisms. Concurrently, there is a discernible increase in focus on comprehensive, 360-degree customer engagement strategies, covering the entire customer lifecycle from initial sales interactions to post-purchase advocacy.
Industry-specific AI transformation is already underway
The transition from AI experimentation to large-scale implementation is manifesting distinctly across various sectors. In the retail industry, AI is revolutionizing store operations through the strategic deployment of cameras and robotics for targeted in-aisle marketing precisely at the point of purchase decision. Furthermore, cashier-less checkout systems are supplanting traditional point-of-sale infrastructure, enabling the redeployment of freed-up human capital to critical back-office and merchandising functions. Within the healthcare sector, Nutanix collaborates with clients on a diverse range of applications encompassing diagnostics, treatment planning, remote patient monitoring, and hospital management, often in conjunction with cloud partners like AWS and Azure. The manufacturing and logistics industries are also undergoing equally profound transformations driven by AI.
The operational challenges of scaling enterprise AI
As AI applications achieve broader adoption, enterprises are encountering a new set of operational complexities. Key concerns for both IT and business leaders now revolve around managing numerous AI workloads and agents, orchestrating infrastructure access across various teams, ensuring robust security and compliance, and seamlessly integrating AI systems with existing business processes. The inherent tension between AI developers striving for rapid deployment and access, and infrastructure teams tasked with maintaining security, uptime, and governance, represents a defining challenge of the current technological landscape. “We are now running multiple agents, all competing for resource access to resolve issues,” Cornely observed. “The critical requirement is infrastructure that allows for the precise definition and enforcement of constraints and resource governance.”
The AI factory: a shared platform for production AI
These mounting challenges are fueling the demand for what Maner and Cornely term the “AI factory.” This concept refers to a unified infrastructure environment designed to support a multitude of users and workloads concurrently, thereby facilitating both experimentation and production while striking a balance between developer agility and enterprise-level governance. At GTC 2026, Nutanix unveiled its Nutanix Agentic AI Solution, a comprehensive platform encompassing core infrastructure, Kubernetes-based container services operating on a topology-aware hypervisor, and advanced services dedicated to the development and governance of AI agents. “We are launching a complete platform, from the foundational infrastructure through Platform-as-a-Service (PaaS) and advanced PaaS offerings, all integrated within a management framework for your AI factories,” Cornely elaborated. “Our goal is to empower teams within the enterprise to achieve genuine self-service capabilities in building these sophisticated applications.”
Hybrid environments are essential to enterprise AI strategy
The effective operation of such an environment hinges on infrastructure flexibility. Hybrid infrastructure is not merely an option but a fundamental necessity. Certain workloads will invariably reside within public cloud environments, while others must remain on-premises due to stringent security mandates, regulatory compliance requirements, data sovereignty concerns, or the need to protect proprietary intellectual property. “Particularly within regulated industries, as data sovereignty and data gravity become increasingly significant considerations, alongside security and the protection of competitive differentiation, the decision will ultimately depend on a company’s specific requirements for its own intellectual property,” Maner explained. This principle underpins Nutanix’s platform strategy. “We provide the ideal synergy, seamlessly connecting applications, data, and comprehensive optimization for these use cases, spanning from on-premises to off-premises and operating in a hybrid mode,” he added. “This capability extends beyond a single cloud, supporting multiple cloud environments.”
This inherent flexibility also extends to the broader technology ecosystem. Nutanix collaborates with major cloud providers such as AWS, Azure, and Google Cloud, as well as regional service providers and emerging “neoclouds.” Nutanix provides these neoclouds with a complete software stack, enabling them to establish and manage their own cloud infrastructures and deliver advanced AI services. This offers enterprise customers already utilizing Nutanix a straightforward extension of their compute, networking, and AI capabilities. Maner characterized this arrangement as mutually beneficial: enterprises gain simplified access to hybrid AI services, while neoclouds leverage a proven, robust platform for their development. Cornely underscored that the entire process is inherently automated and secure by design. “The governance challenges now arising with agentic AI are precisely the same issues we have been adeptly resolving for the past 16 years across all other applications running within your cloud environment,” he stated.
From pilot to production: operationalizing AI across the enterprise
The ultimate objective transcends the successful execution of an AI pilot program; it is about embedding AI into real-world operational use cases. This entails managing infrastructure as a shared, optimized resource, fostering seamless collaboration between infrastructure teams and AI developers, and scaling initial projects to achieve enterprise-wide deployment. “There is a significant disconnect currently between individuals developing AI applications—those AI engineers and agentic AI developers—and traditional infrastructure teams,” Cornely observed. “It is imperative to provide the necessary tooling that empowers infrastructure teams to effectively support AI engineers. This is precisely the value proposition delivered by our agentic AI solution.”
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact [email protected].
Business Style Takeaway: The maturation of AI from experimental phases to large-scale production necessitates robust, flexible infrastructure capable of managing complex workloads and ensuring security. Enterprises must focus on creating “AI factories”—unified platforms that balance developer agility with enterprise governance, particularly within hybrid cloud environments, to unlock AI’s true business value.
Information compiled from materials : venturebeat.com