Witthaya Prasongsin/Getty Images

Data is often called the new oil because of its value to modern business. In the case of ExxonMobil, senior IT executive Andrew Curry says data isn’t just the new oil, but the fuel for a whole range of AI-led initiatives across the company.

As manager of the oil and gas giant’s Central Data Office, Curry is responsible for executing enterprise-wide data principles — and the rapid growth of interest in emerging technologies, such as AI and machine learning (ML), during the past 12 months hasn’t escaped his attention.

Also: If AI is the future of your business, should the CIO be the one in control?

“Everybody’s talking about AI and ML, but you can’t be successful in that area without having high-quality data,” he says, speaking with ZDNET in an interview at the recent Snowflake Summit 2023 in Las Vegas. “The more opportunities that we’re seeing in AI, the more that’s putting pressure on us to ask, ‘Do we have the quality data to be successful in this area?'”

Curry has refined his answer to this question while working for ExxonMobil since 1999. After filling a plethora of roles, and helping the company establish its Central Data Office, he’s learned that one element is crucial to success in emerging technology: data strategy.

Businesses won’t be able to make the most of AI and ML unless they first ensure they have a strong, secure data strategy, which is something Curry has established at ExxonMobil, both during his earlier roles and since moving to his current position as the firm’s data chief in May 2023.

“We understand the quality of our data,” he says. “We know which data is ready for advanced AI and ML capabilities. We also know, and this is equally important, which data isn’t quite there as well in terms of quality.”

Also: AI gold rush makes basic data security hygiene critical

As manager of the Central Data Office, Curry is establishing a long-term data strategy that defines the technologies, processes, skillsets, and rules to help the organization manage its information assets successfully in an age of AI.

As part of its data strategy, ExxonMobil has a technology ecosystem that uses Snowflake’s cloud-based platform. Curry says Snowflake has given the business a consolidated data foundation for the first time.

More generally, his team helps ensure any data initiatives are right for the business. A big part of that effort focuses on the areas where ExxonMobil can differentiate itself through data — and the policies that will allow people to exploit emerging technology in a safe and secure manner.

“You need to know where you can leverage this technology and know where you’re not ready yet,” he says. “We’re really positioning ourselves in the right areas.”

Also: Generative AI is everything, everywhere, all at once

To the end, Curry says his team is evaluating large language models (LLMs) and generative AI capabilities as part of its long-term data strategy.

Right now, there’s not much to talk about in terms of public developments — once again, the focus is on ensuring the firm’s data quality is high before plowing into projects.

“I’ll say there’s nothing in production yet for us,” he says. “But we continue to position that technology and to understand where the business opportunities are, and we’re making sure the data is ready for that work.”

Like so many other executives at large enterprises, Curry is taking a cautious approach to the use of LLMs and generative AIs as part of his company’s data strategy.

Also: Open source is actually the cradle of artificial intelligence. Here’s why

“Generative AI and LLMs is an area where we’ve put a lockdown on it,” he says. “At ExxonMobil, you can’t just go out and use this technology.”

Other research suggests a cautious approach is far from unusual. In fact, many firms aren’t even ready to explore emerging technology.

While 98% of executives polled in a recent survey by Workday believe there are potentially big benefits from deploying AI and ML, almost half (49%) report their organization is unprepared to exploit these rewards due to a lack of tools, skills, and knowledge. 

For Curry, these are significant hurdles that must be overcome before a blue-chip company like ExxonMobil starts experimenting with generative AI and LLMs in a production environment.

Also: AI aims to predict and fix developer coding errors before disaster strikes

“I think there’s a lot of issues about securing your data and thinking about what becomes public information when you leverage some of these public tools, for example,” he says.

“So, we’re taking a cautious approach for broad use within the company, but at same time, there are active plans along the way in a more controlled environment.”

Samantha Searle, director analyst at Gartner, agrees that most blue-chip businesses are taking a careful approach to generative AI.

Like Curry, she says getting your data strategy right is an essential first step for organizations before they start pushing data into LLMs.

“Absolutely, because people aren’t going to like it if you are recommending the wrong things to them,” she said to ZDNET in a video interview. “That’s going to have the adverse effect in terms of customer retention.”

Also: How trusted generative AI can improve the connected customer experience

Searle also points to the importance of data quality. Right now, digital leaders should be concentrating on how their information assets are collected, stored, and exploited.

“It’s very important to make sure that what the AI models predict is accurate. We still need humans to verify the results are right — we know with generative AI that these tools can hallucinate,” she says. “So, you still need verification steps. And making sure you’ve got optimum data quality is a key step to mitigating these risks.”

Back at ExxonMobil, Curry says it’s important to recognize that, while his business is cautious about generative AI, it has a long history of utilizing other areas of ML and AI.

He expects the company to use emerging technologies in several key areas going forward.

First, finance and trading: ExxonMobil has a significant cash flow, and AI and ML could help refine the firm’s processes.

Also: Generative AI tops Gartner’s top 25 emerging technologies for 2023

“Are we exposed in certain areas? Do we have too much cash in certain areas? Should we be investing in other areas? We think we can make improvements on those margins with our cash if we can leverage our ML capabilities.”

Second, supply chain: “Being able to understand when things are being impacted is very important, and being able to react in a timely manner means a lot to the business, and so that is a key area that we’re investing in.”

Finally, Curry points to the potential use of emerging technology at the subsurface level to help interpret trends in seismic surveys.

“That’s an area we’re working to progress,” he says. “It’s probably one area that’s going to have a longer lead time for us as we work to understand it.” 


Please enter your comment!
Please enter your name here