Videos
Data Discipline Is the Real AI Advantage at Total Energies
Insights
- Getting data infrastructure in order is the essential first step before AI can deliver any value; strong systems require investment and planning, not shortcuts.
- Connecting offshore assets to real-time data streams has been a major shift for Total Energies over the past year, enabling AI use cases that were previously out of reach.
- Training domain experts directly on AI tools, rather than relying solely on outside specialists, is proving to be the fastest path to operational efficiency gains.
Most companies treat AI as a model problem. TotalEnergies has learned it's a data problem. Namita Shah, President, OneTech at TotalEnergies, discusses how AI is transforming operations, workforce practices, and decision-making at one of the world's largest energy companies. Shah's approach, centralizing AI search across a 100,000-person workforce, connecting offshore platforms to real-time data, and training thousands of engineers directly on AI tools rather than relying solely on outside specialists, reflects a broader argument she makes throughout: sophisticated AI is worthless without reliable data underneath it. The conversation covers the infrastructure investment required to bring offshore assets online, how predictive tools are reshaping control-room work, and why TotalEnergies' global footprint, spanning Nigeria to Argentina to Malaysia, has made AI adoption a unifying force among engineers regardless of location.
Namita Shah:
Don't underestimate the need to have the right systems in place to get your data in order. It's not rocket science, but it requires investment and it requires planning. Because without your own data, AI is totally useless.
Namita Shah:
My name is Namita Shah. I am in the executive committee of TotalEnergies, which is an energy company. And I'm the head of an organization in TotalEnergies called OneTech, which puts together all the engineers, the researchers, and all the digital of the company in one organization.
Namita Shah:
When we started the AI journey, it was on things like everyday things, searching our own internal technical documents for information because we have a lot of documents, specifications, technical specifications. We have things like safety issues that are centralized, return on experiences, mistakes you shouldn't make before. Those databases were available, and so AI helped a lot to be able to get all the employees of the company, no matter where they work, to be able to just type in and say, I've had this safety problem. What should I do? Which was something that was very difficult for us before, and we have a centralized system on which we maybe have of 100,000 employees, about 25,000 who use that now on a regular basis. So that's a very concrete example of just information sharing. That was much harder to do before we had AI.
Namita Shah:
And now we're going into things that are more complex. So our goal is to be able to use AI for operational efficiency, whether it's optimizing preventive maintenance for all of our machinery and equipment. But we have lots of refineries, we have lots of offshore platforms. To be able to compare how the equipment works between different sites centrally will be a big advantage to try to anticipate when a piece of equipment is going to fail rather than it just failing and then having to deal with it, which impacts production, for example.
Namita Shah:
The scaling AI was really something that was a challenge for us, mainly in the exploration and production businesses because as I said, our assets are all over the world, and they are also offshore. There are very few that are onshore, and when you're offshore you have no connection. To use AI, you need data. To get data, you need to have a connection, which is going to bring it up into the cloud. And that is a big change that we've witnessed in the past 12 months. It's easier said than done because you have to upgrade a lot of telecom systems on all of your offshore assets. But now, we are doing, making the investments to do that. And then the other big thing is we're getting a lot of real-time data off of our assets because without real-time data, you can't really use AI because you need a lot of data to train AI to be able to do everything that I was saying.
Namita Shah:
Geopolitically, I know the world is very complicated. There's no oil and gas in France, so we're not like other oil and gas companies, especially the American ones, who actually have assets at home. We've always had to hedge our bets, spread our risk. We've had assets in a lot of countries all over the world, and when one country is suffering, we hopefully have other countries that sort of make up for that. And what's been interesting is actually AI is something that it doesn't matter which country that we operate in. When we go to the operators on the ground, the young people, it doesn't matter where they are, Nigeria, Angola, Congo, Argentina, Malaysia. It really doesn't matter. When we come with the program and we explain what we're doing, people are on it, really fast. They're very excited about being able to use it.
Namita Shah:
Well, the first thing is when we talk about operational efficiency, what's happening is we are going into areas where our people are not able to do that kind of work. You can't compare millions and millions sets of data as an operator in a control room. So they know that that kind of AI is not something that's there to replace them. It's something to help them monitor better. They do have to change the way they work though.
Namita Shah:
Because when you start having a lot of screens which are sort of flashing, and saying, you know, we have IoT which tracks methane emissions, for example, which is a very important part by the way of our AI program. So you need an operator to say, oh, I see something, there's some methane emissions. I need to go out on the platform. I need to find it. I need to fix it. When you have something which flashes which says your piece of equipment is going to probably break down in two months, then the operator needs to sort of look at that and say, oh, what are we going to do about it? So it's a whole reorganization of the way that they work because they're going to have to catch the information that they get and then relay that on the ground and somebody has to physically then go out into the platform, check the problem, fix the problem. So that's one of the big changes.
Namita Shah:
The other is more profound, which is we have a lot of engineers who really know our domains. I mean, the drillers have a very specific knowledge. Our geoscientists have a very specific knowledge. Our process engineers have a very specific knowledge. And when we want to use AI for these complicated processes, it's hard to find people outside the business who know it as well as we do. So we have to have combined teams between our people and then people outside the business who know AI and know the science behind it.
Namita Shah:
But what we're now discovering is with the new tools that have come out, and this is maybe in the last three, four, five months. The tools are so powerful that if we train our own people on these tools, they're actually able to then connect the knowledge that they have with the tools to develop algorithms that they can use to improve operational efficiency. So our big thing now is how are we going to get all of our engineers, and in OneTech we have 3,400 of them, trained on high-level AI tools so that they themselves can start transforming their own work. Some people are worried about it, but there are others who are very excited about it. So I think we're going to have a lot of ambassadors. We're going to expose them a lot and we've a lot of reassurance that again, we're not doing this to get rid of them. We're doing this because we just want them to be up to speed with what's happening in the modern world. And they shouldn't feel that they're left behind.
Namita Shah:
Well, I would say maybe two things. The first is don't underestimate the need to have the right systems in place to get your data in order. It's not rocket science, but it requires investment and it requires planning. Because without your own data, AI is totally useless.
Namita Shah:
And then the second is take your time to think about what you're going to do because the world changes so fast outside that if you make a decision, either make a decision and execute it like super fast or take your time to make a decision. You're never going to be too late because in six months there's probably a better and newer version that has come out. And once you've decided, stick to it because you're too tempted to use the newest stuff that's come out. But if you do that, you're never going to execute anything within the company. So think, decide, and then when you've decided, act fast, but don't waver too much because you're worried that you've missed the newest thing on the market.
Namita Shah:
Well, we've been trying for a long time to build something with Infosys. And I think that the one thing that has been very useful for me and my teams is to be able to talk a lot to people in Infosys to understand everything that I've said. I think this journey of trying to understand what it means, what you really should be using AI for, what you shouldn't be, what are things that look good but don't bring that much added value. Those kinds of conversations because a company like Infosys has done it for so many other companies. They're able to give us a very honest point of view on what are the real game changers or the things that are really really going to bring value to the company.
Namita Shah:
And I think it sort of opened the eyes of a lot of our engineers in visiting the Infosys campus and things like that as to what's happening outside Europe. Being able to pull them out and having companies like Infosys who are willing to host us and explain what's going on and show us the people on the ground who are doing this in countries like India definitely helps everybody to think about it in a different way.
