VMware executives explain how governance, sovereign infrastructure, security and intelligent software are reshaping enterprise AI adoption.
Pradeep Nair, Vice President, India VMware by Broadcom
As enterprises move beyond cloud-first strategies into an AI-first era, infrastructure decisions are being reshaped by new priorities around cost, data sovereignty, governance and operational efficiency. AI inference workloads, tightening regulatory requirements and the growing need for resilient digital infrastructure are driving organisations to reassess where and how mission-critical applications should run.
In this exclusive interview, Pradeep Nair, Vice President, India and Dilpreet Bindra, Senior Director, Engineering, VMware by Broadcom discuss why private cloud is emerging as the preferred foundation for enterprise AI, how organisations can modernise legacy infrastructure without disruptive rip-and-replace strategies, and why governance, platform engineering and business-IT collaboration will be critical to scaling AI securely and sustainably.
CIO&Leader: Your recent Private Cloud Outlook Report revealed a major shift in workload placement. Production AI workloads on public cloud are declining, while private cloud adoption has increased significantly. Are Indian enterprises leading this cloud reset? What economic factors are driving this shift?
Pradeep Nair: Our survey covered respondents globally, including a significant number from India, and we are seeing the same trend here.
Last year, customer preferences were relatively balanced between public and private cloud. This year, however, we have reached a tipping point, particularly for AI workloads.
Initially, most AI deployments focused on model training, which naturally suited public cloud environments. As AI adoption matures, enterprises are moving towards inference workloads, where AI delivers business value in production environments. These workloads are increasingly shifting on-premises.
There are several reasons for this. Cost is a major factor. Enterprises want greater control over infrastructure spending, particularly as AI inference workloads become larger and more persistent. Running these workloads in private environments provides better cost predictability and operational control.
Another driver is the growing importance of data sovereignty and regulatory compliance. Geopolitical changes have increased concerns around where enterprise data resides. Regulations such as India’s Digital Personal Data Protection (DPDP) Act are making compliance an important consideration, particularly for industries such as banking and telecommunications.
The third factor is operational simplicity. Many AI workloads today are container-based. Enterprises can extend existing Kubernetes and container management skills instead of building entirely new teams. This reduces complexity while accelerating adoption.
Taken together, these factors are driving a significant shift towards private cloud deployments for enterprise AI, and we are seeing this trend very clearly in India.
CIO&Leader: As enterprises move from generative AI to autonomous agentic workflows, technology leaders must balance innovation with security, compliance and profitability. How should Indian enterprises approach this challenge?
Pradeep Nair: Cost management, governance, security and compliance have always mattered. AI hasn’t introduced these challenges, but it has dramatically increased the consequences of getting them wrong.
AI hasn’t created new challenges around governance and compliance—it has dramatically increased the cost of getting them wrong.
When organisations deploy AI without adequate governance, the financial, regulatory and operational risks become significantly larger. That is why enterprises need stronger governance frameworks, tighter cost controls and appropriate guardrails while continuing to innovate.
Business leaders understandably want to launch products faster and stay ahead of competitors. At the same time, they expect technology leaders to ensure that innovation remains secure, compliant and commercially viable.
Technology teams therefore play a critical role in enabling rapid innovation without allowing security, compliance or operational risks to undermine business objectives.
Dilpreet Bindra: AI is fundamentally different from previous technology transitions.
When technologies such as Kubernetes first emerged, development teams could independently deploy clusters and experiment. AI cannot be approached in the same way because the risks are significantly higher.
AI is fundamentally different from previous technology shifts. Without intrinsic security and governance, autonomous agents can create significant operational risks.
Organisations need a carefully designed AI platform with security embedded at every layer. Agentic AI systems require capabilities such as sandboxing, policy enforcement and governance from the outset.
Choosing the right platform becomes critical because it allows organisations to innovate rapidly without compromising security. Without these foundational controls, autonomous AI agents could perform actions across enterprise environments that create substantial operational risks.
CIO&Leader: India’s Global Capability Centres have evolved from execution hubs into innovation centres. How are GCCs leveraging sovereign and private cloud infrastructure to build high-value AI capabilities?
Pradeep Nair: Our experience is that most GCCs continue to work closely with their global headquarters. Their current focus remains more on execution than defining global AI strategy.
Where we are seeing remarkable innovation is among Indian enterprises themselves. Indian banks, government organisations, global system integrators and large domestic enterprises are pushing AI into production at an impressive pace.
We recently hosted an event where one financial institution shared that it already has over forty AI use cases in production. These organisations are moving well beyond pilot projects.
India is adopting AI rapidly, and we are seeing local enterprises combining innovation with strong governance frameworks to deliver meaningful business outcomes.
CIO&Leader: Enterprises are facing infrastructure constraints, particularly around memory and server capacity. How can intelligent software help organisations maximise existing infrastructure instead of relying on expensive hardware expansion?
Dilpreet Bindra: This is precisely where VMware’s strength lies. With VMware Cloud Foundation 9.1, we introduced memory tiering capabilities that extend available memory using local NVMe storage.
Applications continue to experience a large memory footprint while frequently accessed data remains in DRAM and less active memory pages are stored on high-speed NVMe devices. Since NVMe latency is extremely low, application performance remains largely unaffected.
This enables organisations to significantly reduce expensive DRAM investments while extracting greater value from existing infrastructure.
CIO&Leader: Most Indian enterprises continue to operate large legacy environments. What engineering challenges arise when integrating AI into these traditional systems?
Dilpreet Bindra: Enterprises do not need to replace their existing infrastructure.
However, when AI models begin interacting with legacy applications, organisations must ensure their existing data platforms can support increased query volumes and concurrent AI sessions.
You don’t have to rip and replace your infrastructure. Enterprises can modernise incrementally while protecting previous investments.
The challenge becomes even greater with autonomous AI agents interacting across multiple enterprise systems. Existing applications need to become AI-aware, support appropriate identity controls and scale to accommodate these new interaction patterns.
Modernisation therefore becomes an incremental journey where organisations strengthen existing platforms while gradually introducing AI capabilities.
Pradeep Nair: Many of the latest VMware capabilities can be deployed on existing certified infrastructure.
For example, servers already certified for vSphere 8 generally support VMware Cloud Foundation 9. This allows customers to benefit from innovations such as memory tiering without immediately replacing their hardware.
Enterprises can therefore modernise progressively while protecting previous infrastructure investments.
CIO&Leader: As sovereign AI becomes increasingly important, how can organisations implement continuous runtime compliance without slowing application development?
Dilpreet Bindra: Security has to be implemented across every layer of the technology stack.
Data protection, role-based access control, governance policies and least-privilege access should be built directly into AI systems rather than treated as separate controls.
When AI systems interact with enterprise data, organisations need strong policy enforcement to ensure data is accessed only by authorised users and AI agents. These governance mechanisms allow developers to innovate while maintaining continuous compliance.
CIO&Leader: Looking ahead, how do you see the relationship between business leaders and technology teams evolving as organisations pursue sustainable AI?
Pradeep Nair: Traditionally, business teams identified a problem and asked IT to implement a solution.
AI is fundamentally changing that relationship.
Technology teams now introduce entirely new capabilities that business leaders may never have previously imagined. Innovation increasingly becomes a collaborative process where business and technology jointly define opportunities.
As organisations innovate more rapidly, technology leaders must simultaneously ensure governance, compliance, security and operational resilience.
The relationship has become far more collaborative. Business and technology are no longer separate functions—they are becoming inseparable partners in driving transformation.
Dilpreet Bindra: I completely agree. Technology leaders, particularly CIOs, are playing a much larger strategic role than before.
Previously, IT largely responded to business requests. Today, CIOs are helping shape how organisations evolve by introducing new technology capabilities while ensuring they remain secure, governed and aligned with long-term business objectives.
That partnership will only become stronger as AI adoption accelerates.