sdecoret -

CIO interview: Sandy Venugopal, SentinelOne

We speak to the former CIO of Uber and LinkedIn, current CIO of SentinelOne, about how artificial intelligence should be deployed in business

Prior to joining SentinelOne last year, Sandy Venugopal was one of its customers. She was previously CIO at Uber, and before that, was CIO at LinkedIn. Having worked at two companies that are regarded as industry disruptors in the business-to-consumer space, Venugopal is now working at a tech firm that operates in the business-to-business market, providing a platform for endpoint cyber security.

“I’ve used SentinelOne products quite extensively, and I’ve been a huge fan of its product development, design and implementation approaches, and the way it thinks about cyber security,” she says.

It is becoming increasingly common for people in technology to move to the supply side, building the software and systems that support enterprise users. In fact, research from Gartner reveals that IT people are finding more opportunities in tech businesses than in traditional enterprises.

It is easy to see why investments in IT and tech talent benefit tech firms – these investments directly contribute to the company’s revenue. But in businesses outside the tech sector, it is harder to correlate IT investments and hiring tech talent with revenue. Tech is often a cost centre.

Venugopal believes that quantifying tech investment in enterprise businesses needs an objective approach. “Historically,” she says, “even at LinkedIn and Uber, it was all about engagement, such as, How many members do we have?” While this figure may continue to grow, Venugopal says it does not necessarily have a direct impact on revenue or profitability, so sustainable profitability is an important consideration.

Investing in AI

SentinelOne is a customer of AI-powered enterprise search and knowledge discovery tool Glean. In a recent survey of 224 IT leaders at companies with annual revenue of at least $100m and 1,000 employees, Glean found that the budget for generative artificial intelligence (GenAI) projects was set to triple by 2025. Over half (53%) of respondents expected generative AI to pull budget away from other IT initiatives, and 52% also disagreed with the notion that GenAI is a distraction from more important IT activities.

In October 2023, Computer Weekly spoke to Glean CEO Arvind Jain
about linking AI to business intelligence.
Listen to the podcast here:

Venugopal says her approach to investing in AI and GenAI tools is to focus on how the software can help people in the business. “We wanted to be smart about not trying AI in every single area of the company,” she says. Instead, her approach has been to look at a portfolio of business areas in which to try AI. 

“We start with helping our people to make them more efficient and effective, looking at the individual worker and productivity use case,” she says.

This is followed by working with other companies that have products that tackle these use cases in the enterprise, which leads to discussions on current tools like Microsoft Copilot, Google Duet and Glean. “I’m very familiar with Glean,” she adds, “as we also deployed it at Uber.”

Beyond the productivity use cases, Venugopal has also looked at the AI add-ons that are being built into existing enterprise software platforms. “Salesforce and Slack are starting to roll out some AI capabilities,” she says.

Build or buy

As a tech firm, SentinelOne is building AI into the technology it offers commercially. Venugopal sees building up internal AI expertise in product development as strategic. But developing AI for improving productivity and business efficiency is something she believes can be provided by commercially available AI products and services. 

“I’d rather have the few AI expert engineers we can find working on building the product we’re selling,” she says.

Phase one of Venugopal’s strategy has been to focus on efficiency through AI. However, she recognises the true power of AI is in its ability to disrupt industries. She says this phase of a company’s AI strategy may involve re-evaluating and redesigning end-to-end business processes in a way that takes into account how the capabilities of AI will evolve.

“There are going to be some roles that will no longer be needed, or will be done very differently going forward, [due to AI]. We need to be ready to acknowledge and help people in those affected areas to be ready for the change that’s coming”

Sandy Venugopal, SentinelOne

But AI is all about data. “Data maturity is a good way to think about generative AI,” she says.

Venugopal recommends IT leaders assess the maturity of their processes around data and data management and work towards simplifying them. “It’s easy to be more data-driven. Even if it seems people want to get more data, that doesn’t help unless it is well-governed and well-managed data.” 

In Venugopal’s experience, an effective AI strategy starts with having good underlying data. This, she says, is a core component, which provides the right environment for training AI models.

Businesses are at various stages of maturity in terms of their grasp of data. Sound data management is the solid foundation on which an AI strategy needs to be built, as Venugopal explains: “I don’t think it’s realistic to say, ‘don’t look at AI at all until you have a solid data foundation with a very solid data strategy platform and governance’. I think that AI and the maturity of an enterprise data strategy have to run in parallel.”

Venugopal recommends IT leaders focus their AI efforts on areas where they think they have reasonably good data.

Efficiency gains

From her own experience, it looks like there is certainly a case for CIOs and business leaders to think of AI as a way to help the business become more efficient. For instance, some tasks people do are ripe for automation, and some of these may be infused with AI and become hyper-automated. While the tech sector positions the current raft of AI tools as a way to augment human workers, as a task becomes more automated, the number of hours it takes someone to finish it is reduced.

As hyper-automation is scaled up across an organisation, it is obvious that less work time will be needed by human workers since, under the guise of augmentation, the AI is removing the arduous manual tasks.

“There are going to be some roles that will no longer be needed, or will be done very differently going forward,” says Venugopal. “I think we need to be ready to acknowledge and help people in those affected areas to be ready for the change that’s coming. That’s part of our responsibility.”

Read more about AI deployments

Read more on CW500 and IT leadership skills

Data Center
Data Management