Can you unpack the ‘HPE is hybrid by design’ philosophy for the UAE audience?
Gilles: At HPE, we identified very early that hybrid cloud would become the de facto operating model for enterprises and the public sector. That was our North Star at a time when not many were talking about hybrid the way we were. Fast forward to today, and that vision is now a reality.
The challenge we see with most customers is that they’ve ended up in what we call “hybrid cloud by accident”. They moved some workloads to Azure, individual business units pushed others to different public clouds, and they still have legacy infrastructure on-prem alongside private cloud setups. The result? Unmanageable complexity. That’s the number one concern we hear from customers.
So, ‘hybrid by design’ is about taking a step back and being intentional. For us, cloud is not a destination; it’s an experience. It means having the same ease and elasticity regardless of where your workload lives.
That’s why we built the HPE GreenLake Cloud Platform, which delivers that consistency across the entire stack – networking, compute, storage, and private cloud.
We’ve also made two strategic acquisitions to help customers manage the hybrid environment. OpsRamp gives you a single pane of glass across your entire estate with AI-powered observability and troubleshooting. And Morpheus lets you automate workload provisioning across any environment, public or private, through one consistent tool. That brings real simplicity and cost efficiency.
The philosophy is simple: we want customers in the UAE and across the region to move from complexity to control – hybrid cloud, by design.
How does it specifically address the tension between innovation and the strict data residency requirements in regulated sectors like government and finance?
Gilles: A trend that has really accelerated over the last three to four years is sovereignty, and it’s very relevant here in the UAE, across the EU, and in regulated markets globally. The core question customers are asking is: “Where is my data, and who controls it?”
What people want is more control over their data, stronger security, and at the same time, they don’t want to sacrifice the cloud experience. AI is probably the most powerful example of this tension playing out in real time. We’re seeing customers start their AI journey with experimentation in the public cloud, which is fine for early-stage work. But when they need to scale AI into production, they move it on-prem. The reasons are consistent: they don’t want their sensitive data leaving their environment, they have regulatory obligations, and frankly, the cost economics shift significantly at scale.
To address this specifically, we co-developed with NVIDIA a private cloud offering purpose-built for AI, a fully pre-integrated solution that brings the AI factory experience on-premises. So, for a government entity or a financial institution in the UAE that cannot move its data to a public cloud, they don’t have to compromise. They get a production-grade AI environment, within their own walls, with full data sovereignty intact.
Innovation and compliance don’t have to be in conflict.
What separates basic AI adoption from the “AI factories” you’re powering locally?
Gilles: Studies suggest that between 80 to 90% of AI proofs-of-concept don’t make it to production. We’re seeing that trend shift, but currently, there are two distinct challenges holding customers back from moving ahead with AI. The two root causes are consistent: unclear use cases and the complexity of building the data foundation and infrastructure needed to actually run AI at scale.
Building an AI factory inside an enterprise, especially for the 80% of the market that isn’t a hyperscaler or a highly specialised tech organisation, is genuinely hard. Most of these businesses don’t have deep AI infrastructure skills in-house, and integrating the data layer on top of that adds another layer of complexity entirely.
That’s exactly the gap our Private Cloud AI is designed to close. We’ve done the integration work, the testing, the validation, so our customers don’t have to. What would typically take six months to stand up, a fully functional AI inferencing environment inside an enterprise, can now be done in a matter of weeks.
For a business in the UAE looking to move from experimentation to real, production-grade AI, that difference in time-to-value and cost is significant. The AI factory concept isn’t just about hardware; it’s about removing the complexity that’s been the single biggest reason AI projects stall between pilot and production.
UAE is advancing towards smart cities, and we have massive projects as well for that. So why do you think edge processing is critical here compared to markets relying on the centralised public clouds? And how does HPE’s hybrid approach enable real-time decision making?
Ahmed: The UAE has genuinely been a pioneer in Smart Cities, and HPE is proud to have been a key technology partner in landmark projects. The boldness this country has shown in adopting and scaling new technologies has always been remarkable.
Recently, UAE officials indicated that the direction is clear, with a commitment to what they call the “ruthless adoption” of agentic AI, and ambitions to deploy AI agents at a truly national scale. When you consider what it takes to deliver that at scale, latency becomes the defining challenge. Agentic AI applications cannot reliably serve millions of end users through a centralised public cloud model without encountering performance bottlenecks.
Edge processing becomes critical here. While deploying AI-driven services that need to touch end users in real time, across a smart city, government services, or connected infrastructure, the compute needs to be closer to where the action is.
And underpinning all of it is networking. Connectivity, low latency, security is genuinely underestimated in most conversations about AI adoption. Connecting everything together without bottlenecks and doing it securely at scale is the real infrastructure challenge.
This is precisely where HPE’s acquisition of Juniper becomes highly relevant for the UAE. It gives us edge capabilities and networking capabilities that are purpose-built for exactly this kind of large-scale, latency-sensitive deployment.
HPE has been working on liquid cooling technology for decades. That’s not a talking point; it’s decades of intellectual property that directly addresses one of the most pressing infrastructure challenges that comes with deploying AI at a national scale. That depth of experience positions us well for whatever the next wave of challenges looks like here.
Any custom wins or pilots which are demonstrating hybrid cloud, closing the execution gap for governance initiatives?
Ahmed: There are two references in the UAE that we’re particularly proud of and that I think speak directly to our ability to deliver real hybrid transformation, not just in concept, but in execution.
The first is the Commercial Bank of Dubai. This is a meaningful example of the kind of transformation we’ve driven in the financial sector, an industry where the stakes around data, security, and continuity are as high as they get.
The second is Emirates Global Aluminium. What makes this one stand out is that it didn’t just resonate locally; it captured global attention for the sheer uniqueness and scale of the transformation. We played a central role in it, and I think it’s a compelling case study in what hybrid cloud delivery looks like when it’s done at an industrial scale.
These two engagements are a testament to what HPE can deliver in this market, across financial services and heavy industry, two very different sectors with very different demands. And in both cases, the hybrid cloud wasn’t a theoretical framework. It closed an actual execution gap for organisations operating at a significant scale in the UAE.
How does HPE tailor the hybrid solutions for the for this region’s unique blend of public sector ambition and private sector agility?
Gilles: This region is genuinely at the forefront of the thinking around sovereignty, cloud strategy, and technology investment, and that’s something we pay close attention to, because what works here often becomes a blueprint for other markets globally.
On sovereignty specifically, one capability that’s particularly relevant for the UAE’s public sector and sensitive industries is what we call our air-gapped solution. It’s a special instance of the HPE GreenLake Cloud Platform that is completely disconnected from the internet, so you get the full platform experience, all the cloud capabilities, but in a totally secured, isolated environment. That’s a hard requirement in defence, certain government functions, and other sensitive sectors. And frankly, the demand we see for this in the UAE mirrors what we’re seeing in the EU and other sovereignty-conscious markets around the world. The requirements are more similar than people might expect.
The other major theme we’re seeing, and it’s as significant for many customers as AI, is infrastructure modernisation. That shift has fundamentally changed the virtualisation landscape for many organisations, and there’s cost pressure that comes with it. We’re hearing from customers that rising virtualisation costs are actually limiting their ability to invest in new priorities like AI.
Our response to that is HPE Morpheus VM Essentials, a cost-effective, agile alternative to traditional virtualisation that gives customers a credible path forward without the financial burden. It’s something we’ve made a strategic decision to lean into, both for our customers and our partners, because the need is real and urgent.
So tailoring for this region is about combining sovereign-grade security with the commercial agility to modernise infrastructure and free up investment for the next wave of innovation.
For enterprises hesitant about hybrid due to cost or complexity, what’s the clearest ROI story from your recent UAE deployments?
Gilles: Most aren’t hesitant about the direction; they’re asking how to make the financials work. We start with a Cloud Physics Assessment: a full analysis of where workloads sit, how infrastructure is utilised, and what it’s costing. More often than not, we find infrastructure running at 60 to 70% capacity, with ageing equipment driving disproportionate energy costs.
We come out with a concrete recommendation and a business case, not just a technology conversation, but a financial one. And we have a Solution Centre here in the UAE, one of a handful globally, with in-country technologists to deliver that white-glove experience at scale. Global capability, genuine local depth, that’s what makes the ROI story real.
What should UAE CIOs prioritise in hybrid infrastructure to future-proof against evolving AI demands and sovereignty regulations?
Ahmed: Readiness. True hybrid capability gives a CIO the diversity to provision workloads wherever makes sense. The pressure from boards and C-suites to move fast on AI is universal, but the answer isn’t to rush in without the right foundation.
Start at a scale that’s truly expandable. Our private cloud for an AI appliance, co-designed with NVIDIA, is plug-and-play, scalable from a minimum viable capacity upward. It condenses what is genuinely complex infrastructure into something deployable without months of engineering effort. These appliances are already live in the UAE, powering services used by millions. That’s not a pilot, that’s production.
Get the hybrid foundation right. Build sovereignty and security from day one. And adopt an AI infrastructure that lets you start small and scale fast. The technology is ready. The question is whether your infrastructure is.
Discussion about this post