Videos
Databricks on AI, Data Platforms & Energy Transformation
Insights
- The biggest AI opportunities in energy remain grounded in practical business outcomes, including production optimization, predictive maintenance, root cause analysis, and energy trading.
- Successful AI adoption depends on trusted, governed, and unified data foundations that enable organizations to scale AI beyond isolated use cases.
- Data platforms are becoming the backbone of enterprise AI by enabling agents and models to reason across the full breadth of organizational data.
At the AI Horizon event in Houston, Sandeep Garg, Vice President Infosys Energy Practice, speaks with Julien Debard, Global Director, Energy and Utilities at Databricks, about how energy organizations are moving from AI experimentation toward scalable business outcomes. The discussion highlights continued demand for machine learning applications such as predictive maintenance and root cause analysis alongside emerging opportunities in generative AI and autonomous agents. Julien Debard explains why production optimization and energy trading are becoming high-value AI use cases, where even small improvements can generate significant financial returns over long asset lifecycles. The conversation also explores the growing importance of unified data platforms that bring together operational, market, and telemetry data, enabling enterprises to govern information effectively and deploy AI models and agents at scale. Together, they outline how the Infosys–Databricks partnership is helping energy companies accelerate value realization by combining industry expertise, modern data architectures, and AI-driven innovation.
Sandeep Garg:
Hi, I'm Sandeep Garg, Vice President Infosys Energy Practice.
Julien Debard:
I am Julien Debard, I lead the energy and utilities at Databricks.
Sandeep Garg:
Welcome Julien to our AI Horizons Day. What's your take on what you heard from these experts on the joint initiatives that we are running on production optimization and energy trading space?
Julien Debard:
Yeah, a lot of excitement and thanks for putting such a day together with a packed room and some very good panels. So thanks for doing that.
Julien Debard:
Look, interestingly enough, the discussions that we've had on the side of the event were around the traditional AI, the machine learning. People are still very interested in what we can do in terms of predictive maintenance and root cause analysis and this kind of thing. So I find that very interesting. Obviously, we talked about the more fancy GenAI, chatbot, passing of documents and agents. But very, very down to earth, people understand that there is a ton of value still to be made with the machine learning and AI space. And then, as it was presented in the different panels, I was also surprised that our customers really understand that in order to do AI you need good data. You need a unified data layer that is orchestrated and governed so you know what you are working on and you can trust the data that you're working on. So very interested to see how educated people are, both on the need for that data but also excitingly for me, what Databricks does in that space.
Julien Debard:
Look, we talked about the solution that we're coming up with together around the production optimization and energy trading. I think production optimization people understand, production lasts 25 or 30 years. So obviously optimizing just a percent of your production can turn into billions of dollars of benefits, but also production evolves over time. So people understand that AI will be the one that can adapt as your production changes. So very excited to have the discussion with our customers on how we can do that.
Julien Debard:
The energy trading is one that is a bit more complex. I think again, interestingly enough, people understand the value of our platform not just as a data platform but as a data and AI platform now. And really understanding that when you want to do energy trading, you need to understand all of your supply, you need to understand all of your demand, and you need to understand how things flow between each other. So obviously on our platform we are able to bring all the telemetry for your power generation or your production of oil and gas and really understand what you're producing today, what you're forecasting to produce tomorrow. So then naturally, when you have all of that on one platform, you want to do trading because you are very smart about, that's the kind of things I'm going to be able to sell very soon in the future and that's what I can be buying at a certain price today. So I think a ton of excitement. I'm very excited about the energy trading space as well. So I think our customers like that we are working together in that space.
Sandeep Garg:
And I think that's where our partnership comes in, which is to bring a partner like Infosys, they're going to be able to first guide you through all the steps that are required to get there, but also help you start small so you deliver value quickly and then you grow your use cases as you bring more data sets and you understand more of the AI tools. So super exciting day.
Julien Debard:
Great thoughts Julien. I think I'm also very excited, specifically on these two. The industry needs these solutions. So that's something which I'm very excited about and glad to be partnering with Databricks on such cool solutions that we will take to the market. Now, just again a follow up on that, how do you think the platform approach can accelerate this whole build phase that we are trying to do and bring these solutions to the market faster? How does the platform approach help us?
Julien Debard:
So I think the whole platform vision is really becoming clear to everybody. The idea of getting all your data in one place, access to all your data in one place, so you can be combining these different use cases. And so what's really growing now is the opportunity cost between, I have all this data, but I also have all these silos in my organization. How do I combine these two so I make sense? And that's where the platform makes sense.
Julien Debard:
But that platform vision now brings the idea that we don't have to wait for the org chart to catch up with technology, because you can give into the hands of someone the data that they need. That data platform, now they can find it. They have a catalog where they know all the data that they have access to. That's why it is so important to have this data platform. But then obviously, once you have access to all of your data in one single layer and you want to govern who has access to it — the question is where do I run all my agents and when do I run all my AI? It makes a lot of sense to run it right on top of where your data sits, it's a lot more effective. But also now your agents are reasoning on the entire set of data of your company. So you don't have to just point them to a certain amount of data, you can tell them to reason on the entire data that we have in here.
Julien Debard:
So I think a lot of panel discussions were around that, the importance of having a data platform. I'm very excited by the technology that's coming out there. Obviously our data intelligence platform is growing very, very fast. We give that basis where now you can try AI, you can make it happen very quickly. Any new solution, any new model, any new agents, you can very quickly try it at scale because your data is in one place. So as you can notice, I'm very excited about what the future holds. Very excited to do this with Infosys as well.
Sandeep Garg:
Sure, thanks a lot Julien. I think we are excited too. These are some great things that we are working on together. The platform approach is definitely going to have a great boost for the industry, to be able to do things a lot more faster, in a more holistic way as you describe. So very excited to be on the journey and let's hope we get these solutions out to our clients soon. Thank you so much. Appreciate you coming out over here.
Julien Debard:
Thank you very much. Let's solve all these energy challenges together.
Sandeep Garg:
Sure.
