India’s young, tech-savvy workforce is essential in addressing the global AI talent shortage: Jay Upchurch, SAS

In a recent interaction with  Express Computer, Jay Upchurch, Executive Vice President and Chief Information Officer, SAS, shares how SAS is evolving its cloud strategy with AI-driven analytics at the core, the power and promise of its Viya platform, and why India plays a pivotal role in SAS’ global innovation roadmap. From managing rising cloud costs and driving ethical AI adoption to nurturing early-career talent in its R&D hub, Upchurch offers a candid perspective on the opportunities and challenges in today’s data-driven enterprise landscape.

SAS has been a pioneer in analytics, and with AI and cloud computing transforming industries, how is SAS evolving its cloud strategy to stay ahead? Could you share insights into how SAS is integrating AI-driven analytics into its cloud offerings to enhance business decision-making?

First, we’re absolutely focused on helping our customers move safely and confidently to the cloud. Analytics workloads in the cloud are quite different—they’re unpredictable and varied. For example, a data scientist might run a heavy algorithm that consumes massive compute resources, while a business analyst might just be querying data for a simple dashboard. This variability is hard to manage in traditional on-prem environments due to compute limitations. So, enabling a cloud environment that can handle this variety is crucial.

Second, we’re very focused on cost efficiency. There are many stories of CIOs surprised by unexpectedly high cloud bills or debating repatriating workloads. We make sure our initial cloud footprint is cost-effective and work closely with engineering to ensure our software runs as fast and efficiently as possible. In the cloud, time is literally money—the faster your job completes, the sooner you stop the bill.

In fact, we conducted a third-party study comparing our platform, Viya, to both open-source and off-the-shelf competitors. Viya was found to be over 30 times faster. That speed not only delivers insights faster but also saves costs significantly.

Third, we focus heavily on workload management. Inefficiently submitted jobs can drive up cloud costs. We help customers optimise how they submit jobs, taking advantage of lower-cost cloud resources based on region or time, ensuring better, faster, and cheaper performance.

So, whether it’s cloud transition, cost optimisation, or intelligent workload management, we’re evolving our cloud strategy with AI-driven analytics at the core to help businesses make smarter decisions.

Since you mentioned the Viya platform, could you please elaborate on it?

One of the great things about our Viya platform is that it supports the entire AI lifecycle, and it’s structured into three main parts. The first part focuses on data—managing, accessing, and preparing it. The second part covers the data science side—modeling—and the third is about deploying, tracking, and managing those models.

That first part, the data layer, is critical. Where is your data? Can I access it easily? Is it explainable and ready for computation? Our goal is to let the data stay where it is—whether it’s on-prem, in the cloud, or across different platforms. We’re not trying to win the data platform game. Some of our competitors start with a data story and then try to add analytics on top. We approach it differently—we go directly at analytics and AI, while still ensuring your data is modern, accessible, and ready to support that journey.

What trends have you observed in cloud adoption among your customers? With rising cloud costs, some businesses are shifting back to on-premise, while others embrace a hybrid model. How do you see this evolving, and what is your perspective on the way forward?

The reality is that cloud costs aren’t drastically different from running your own data centre—if you’ve already built it. Sure, you can buy a server for less than it costs to rent one in the cloud, but the cloud takes care of everything else, critical facilities, lifecycle management, power, cooling, and maintenance. When you factor in the full cost of running infrastructure yourself, the difference narrows significantly.

Many companies struggle with cloud costs because they’ve already made significant investments in on-prem infrastructure years ago and want to maximise that sunk cost. But there’s a misconception here—cloud isn’t necessarily more expensive. The difference is you see it as a daily or monthly bill, which makes it more visible. On-prem inefficiencies often go unnoticed, but in the cloud, every inefficiency is billed and tracked.

Now, about the trend—CIOs aren’t really pulling workloads back from the cloud. Those who moved early expected lower costs, but now realise cloud’s value lies more in flexibility. You get elasticity—scale up or down as needed. However, without proper FinOps practices, you can overspend.

What we are seeing is growing interest in hybrid models. Customers want a stable on-prem baseline with the ability to burst into the cloud when needed. This makes sense—though it gets trickier with data placement, sovereignty, and compliance, especially in regulated industries. So while hybrid is appealing in theory, it’s a bit harder to execute in practice. Still, it’s the direction things are moving.

 As CIO of SAS, how do you define data intensity, and how can enterprises effectively leverage it to drive digital transformation and business value?

When it comes to data, especially in the context of analytics and AI, I see many customers and peers with big AI ambitions but who haven’t taken the time to properly manage their data. They spend a lot of money chasing advanced goals without laying the proper foundation. To effectively leverage data for AI, you need to start with good data hygiene,  make sure your data is in a modern system, properly catalogued, well-governed, and maintained. Without this, your AI efforts will be costly and inefficient.

Another key aspect is ensuring your staff has strong AI literacy. While it’s easy to play around with AI as a consumer, applying it effectively in an enterprise context, especially with Gen AI, is more complex. We also need to be cautious about “shadow AI”—the risk of data unintentionally leaking into public models if employees aren’t careful. It’s one thing to have an ethical policy, but it’s another to help staff understand the responsible way to approach data handling.

At SAS, curiosity is a core value, and we encourage our employees to explore and innovate. As IT, it’s my job to provide the necessary guardrails to keep our company and customers safe while allowing that curiosity to flourish.

Given the global talent shortage in the industry, how do you view India’s role in addressing this challenge? With a young, tech-savvy workforce, do you see India emerging as a centre of excellence for AI and analytics talent? How is this reflected in SAS?

Yeah, the global talent famine in the industry is very real—there’s no doubt about that. One of the things I love about being in India is the sheer volume of available talent. And when it comes to artificial intelligence, data, and analytics, the younger generation here just gets it—they use it daily, it’s native to them.

Given that the average age in India is around 28, with about 60% of the population under 32, it’s an incredible demographic advantage, especially when you compare it to aging populations in other parts of the world, like Japan.

At SAS, the majority of our employees are under  30 and early in their careers in India. What’s remarkable is they’re coming in with strong foundational knowledge—I’m not having to train them from scratch or push AI literacy. In fact, they’re often the ones pushing us forward with new ideas and curiosity. It’s exciting to see ideation coming from this centre, and that’s a big part of why it’s so energising to be here.

So this Pune site is your R&D centre in India, right?

Yes, it is one of our key R&D centres. In India, SAS has around 1,000 employees, and about 800 of them are based at this Pune site. Roughly 400 to 500 are focused purely on R&D for our core products, while another 300 are part of our IT and cloud operations.

Globally, SAS also has R&D centres in Glasgow, Beijing, and at our headquarters in Cary, North Carolina, along with smaller satellite offices. For me and my division—which covers IT and cloud—this Pune centre serves as the hub. You could call it our GCC. It supports all IT services and also houses our global operations centre, which manages our cloud 24/7.

Since you mentioned that SAS operates as a GCC, do you have any upskilling programs in place? Given the vast talent pool available, what strategies do you use to nurture and retain this talent, particularly in this region?

Want to guess what our attrition rate was last year in my division? It was just 3.2%. Compare that to the industry average in IT across the country—close to 20%. That’s a number I’m incredibly proud of.

So why was it so low? First, we treat early career talent with a clear focus on career progression. When someone joins us straight out of university—often through strong partnerships with local universities—we make sure they understand what their growth path looks like. 

Second, we’re growing. And growth naturally creates new opportunities. So here, progression isn’t based on tenure—it’s based on your capability and the opportunities that come with that. I think that’s a big differentiator.

Third, we invest heavily in nurturing talent through training, education, and certifications. We encourage it, reward it, and set employees up with structured programs to help them succeed.

Ultimately, it all leads to meaningful work. Personally, I want to be part of a company that’s doing great things, surrounded by good people, and where I feel valued. I believe we do that really well—especially here, where a large portion of our workforce is early in their careers.

With AI adoption increasing rapidly, there is growing concern around ethical AI and governance. How are you ensuring responsible AI adoption in your analytics solutions, and what role does IT leadership play in enforcing ethical AI practices?

First, are we doing anything about ethical AI? Absolutely, and I say that proudly. Back in 2021, we established a dedicated Data Ethics Practice led by Reggie Townsend. At the time, there was growing concern around bias, especially given the global social movements highlighting systemic inequities. We started asking ourselves, does our data reflect historical bias? And do people realise that models trained on such data could perpetuate it?

This division was created as a non-commercial initiative, focused solely on ensuring the models we build are transparent, trustworthy, and fair. It played a major role in shaping our internal culture around responsible innovation.

Second, in terms of impact, Reggie and his team have been instrumental globally. He’s part of the U.S. Presidential Advisory Committee on AI and has worked closely with governments in Singapore, the UK, Germany, and more to advise on policy frameworks. The goal has always been to bring some consistency and guidance to how ethical AI is implemented around the world.

This work has also led to real product innovations. For example, we now use “nutrition labels” for models—similar to food labels—detailing what data was used to train the model, who trained it, which algorithms were used, and more. It’s designed to make model decisions explainable and auditable, which is especially crucial in sensitive areas like credit decisioning in banking, where unintentional historical bias can harm certain groups.

Additionally, during the model training phase, especially in our Viya platform, we flag potential bias in real time so data scientists can address it before deployment. It’s all part of our broader mission to drive responsible and transparent AI adoption.

With AI, automation, and analytics reshaping IT leadership, how do you see the role of CIOs evolving in the next five years? What skills and strategies should CIOs prioritise to drive digital innovation effectively?

First off, I love being a CIO—especially today. Ten years ago, I’m not sure I would have enjoyed it as much. Back then, CIOs were expected to be the smartest person in the room, taking orders, building solutions, and delivering them back. We were mostly a back-of-house function. But today, it’s completely different. We’re at the business table, partnering with every function—legal, finance, HR, R&D, consulting—helping them think through how technology can improve products, services, and efficiency. That’s exciting.

My role is about bringing solutions to the table—whether it’s AI, agentic AI, or any emerging tech. I’m not asking business leaders to reinvent what they do. I show them possibilities and ask, “Have you thought about this?” I get to help people work better, faster, cheaper—and ideally, with a better work-life balance. That’s something I truly enjoy.

As for the skills CIOs need today, I’ll share three big ones:

1. Storytelling.
   CIOs now must be exceptional storytellers. It’s not enough to just deliver a solution—we have to tell the story behind it. If I can’t explain the value and outcome of a tech investment to my CEO or the board in an inspiring way, I won’t get the funding. Storytelling is how we drive adoption and support.
2. High emotional intelligence.
   We’re constantly driving change, which means disrupting how people work. Technology alone isn’t enough—change management is crucial. At SAS, for example, we’re preparing for an IPO, which requires rolling out entirely new financial systems. That means shifting long-standing processes for employees who’ve done things the same way for decades. I need to bring them along on that journey, help them embrace the “why,” and guide them through the change.
3. Curiosity and continuous learning.
   Technology is evolving fast—especially with GenAI. At SAS, we’ve been in the AI space since the ’70s, but GenAI has accelerated everything. CIOs must stay curious and explore what’s possible, while also balancing that with delivery. We can’t chase every shiny new thing—we have to stay grounded in our business goals. So, it’s about managing that balance, staying ahead of the curve, but still executing on what matters today.

And finally, I’m less concerned now with deep expertise in any one technology—because it all changes so fast. What matters more is understanding how it applies to the business. At the end of the day, IT exists to serve the business. We have to align our technology, people, and processes with the company’s strategic vision.