Large geological data sets provide the foundation for powerful new applications in the energy industry, thanks to advanced data analytics and artificial intelligence.

But read their press releases and it sounds like data companies are busy “implementing solutions,” “empowering workforces” and “optimizing outcomes.” It’s as though you’re going to call a company and someone will say, “I can’t talk right now, I’m leveraging a synergy!”

Data analytics and AI have gone through a period of maximum hype recently, with a side order of paranoia. Too many people think AI will perform miracles and then take over their jobs … and maybe even the world.

Beyond all the jargon and hyperbole, practical tools and useful results really are coming out of these advances. The energy industry has already seen significant improvement from new applications of Big Data, in almost every area of operations.

Now the data experts have their sights set on another breakthrough – a sort of Holy Grail of practical data application: the Answer Machine.

Geology-Based Performance

Jon Ludwig is president of Novi Labs, an energy analytics company in Austin, Texas. This year it received a $35-million investment from a private equity firm to improve and expand services, including making strategic acquisitions. Novi Labs described the investment firm as “focused on backing category-defining companies leveraging artificial intelligence in high-impact industries.”

Ludwig said Novi Labs concentrates on finding the best allocations of capital in oil and gas. It uses data tools and AI to analyze and forecast production.

“At the end of the day, everything you’re doing depends on the production from the well, because that determines the return on investment,” he said.

Forecasting future output from a well involves a different type of problem from data applications in the internet world, Ludwig noted, calling them “two very different things.”

“Forecasting well production isn’t anything like predicting website clicks. They’re worlds apart. A well’s performance depends on geology, engineering and market forces. The complexity is higher, the stakes are bigger, and the cost of being wrong is measured in millions (of dollars), not clicks,” said Mohamed El Hannaoui, Novi Lab’s vice president of marketing.

Accessing accurate and reliable data, and lots of it, is key to production forecasting. But analysis also requires generating additional, synthetic data, Ludwig observed. AI excels at both analyzing historical datasets and creating synthetic data to fill in any gaps, he said.

In the forecasting process, you need abundant real-world data and “you also have to create a lot of data, as well. It’s sort of a two-part problem,” Ludwig noted.

Other companies are using geodata and data analytics to assess potential production as early as the prospect level.

Wood Mackenzie introduced two AI-powered tools at AAPG’s IMAGE 2025 meeting in August designed to “eliminate human bias from oil field and exploration prospect evaluation.” It said the tools can help predict which reservoirs will succeed before drilling begins.

Specialized Advances

Ludwig cited three areas where a combination of geodata, analytics, machine learning and AI is making a practical contribution to the energy industry:

u Providing production forecasting at scale: “If you think about drilling a well, the first thing you want to do is think of all the ways you can drill that well. AI is very good at that,” Ludwig said.

Operators now can evaluate likely production results in a play and choose the best drilling and completion approaches to meet their objectives, in terms of costs, financial returns and total output.

“AI can evaluate countless options in the time it might take an engineer to fully analyze just one. That’s why companies now run thousands, even millions, of scenarios and choose the path that delivers the best results,” El Hannaoui said.

u Real-time optimization of drilling operations: Ludwig described this as monitoring for efficiency and for “knowing when the drilling is going sideways or really isn’t going as well as it should.”

Data analytics has helped the industry achieve high total production rates even as it has reduced the number of wells needed to attain that production, he noted.

“Historically, if you look at production compared to rig count, there’s your proof,” he said.

u Well production optimization: Advanced data applications have led to better well-siting and spacing, completions improvement, more efficient well operations and also the ability to anticipate downturn for proactive response.

“Enhanced oil recovery is another thing where it’s made a significant difference – waterfloods, workovers, intervention in the well,” Ludwig said.

In data applications, everything starts with the data used, both its quantity and quality, he noted. Like many other companies, Novi Labs highlights its ability to access a full spectrum of data including proprietary data sets, and the quality of its own proprietary data algorithms.

“There’s a lot of data out there, but not all of it is in good shape. AI is good at pulling data together and cleaning it up,” Ludwig observed.

“We can publish actual production data instead of what’s being reported to the state, and we can also use AI to generate data, to create synthetic data. We use AI algorithms against our data and learn what the outcomes are,” he said.

Enter: The Answer Machine

Data companies have already started to introduce AI-based programs that can respond to questions relevant to specific industries, including for the energy industry. Ludwig called that type of application an “Answer Machine.”

“We’re working on building AI advisers to allow users to ask questions in real time and get answers right then. We’re working hard to create that kind of future,” he said.

While those applications are still in their preliminary stages, they are going through a process of continual improvement. Data experts hope to create AI advisers that produce only highest-reliability responses, the ultimate Answer Machines.

“When you’re using ChatGPT, it will always give you an answer. It might not always be right,” Ludwig noted – and similarly, a petroleum engineer or petroleum geologist can always provide an answer, of some kind.

“A petroleum engineer or geologist will always give you an answer, but it’s not always the best one. That’s true for any knowledge worker. With an Answer Machine, they can combine their expertise with AI-driven insights,” El Hannaoui said.

“And when you’re seeking approval for a $150-million drilling program, you want recommendations backed by solid data. The idea is not to reduce the number of engineers. It’s to give them more tools when they’re out there making million-dollar decisions,” he added.

High Stakes Calls for High Reliability

Ludwig said it will take time for the data industry to develop improved, comprehensive, high-reliability applications for the energy industry, and for the energy industry to adapt to new processes and to adopt those new data tools. “High reliability” is an important concept.

“It’s very expensive to drill wells, so the industry doesn’t just jump on the train and take off down the track,” he commented.

Novi Labs started when the combination of advance data analytics, machine leaning and AI was in its infancy, going through a learning and development process along with the industry, Ludwig said.

“We’ve spent 10 years building all of the core stuff. It’s taken time to get all that right – it’s not just throwing an algorithm at some data,” he noted.

The next step involves refining today’s data applications, improving the reliability of output, achieving better results and taking AI and data analytics into new areas.

“I think we have a long way to go to know what’s possible,” Ludwig said.