Inside LVMH's perfectly manicured data estate, where luxury AI agents are taking root

Anthony Cirot
Managing Director Google Cloud France
In an interview as wide-ranging as LVMH's 75 unique brands, CIO Franck Le Moal discusses the data foundations that paved the way for AI perfectly suited to each maison.
How do you maintain the status of the world’s most luxurious brands when, in the digital age, seemingly everything is at our fingertips, just a few keystrokes away?


If you’re the technical tailors at LVMH, the world’s foremost luxury conglomerate, all that data becomes like the sequins on a Dior evening gown or gems on a Tiffany necklace: attractive on their own but truly stunning if you can weave them together and keep your clients coming back for more (especially when those pieces would look so great together).
“Well, even when we step into the digital world, our motto is still to drive to the store as well — digital is not the end of the story,” said Franck Le Moal, chief information officer at LVMH. “But we do need to adapt also the way we are interacting with our customers, to meet them where they are, which is so online these days.”
What most sets LVMH’s 75 distinct and distinguished maisons and brands apart is, as Le Moal notes, “Crucially, you still have a client advisor, and that’s who we want to empower with data and AI.”
For four years now, LVMH has been working with Google Cloud to build a data foundation for those brands, which include its namesakes Louis Vuitton, Moët & Chandon, and Hennessy as well as Dior, Tiffany’s, Bvlgari, Sephora, Celine, Dom Perignon, and so on. As if meeting the needs of luxury brands and their exacting clientele were not already an order taller than the Eiffel tower, the LVMH tech team also had to keep that data separated, differentiated, and secure.
Anthony Cirot, Google Cloud’s vice president form EMEA South, sat down recently with Le Moal to find out how they’ve built and manicured this exquisite data estate — as well as how they were ready just in time for the explosion of generative AI that has seized the luxury industry, and the ways that this, too, presents unique challenges to this one-of-a-kind brands.
Cirot: You kicked off your data journey in a structured way around five years ago. What was the goal, and where did the LVMH team start?
Le Moal: When we got started, there was a strong conviction that we should start leveraging our pieces of data first and foremost to better address our customers. For all the care and attention that we give to our customers in our shops and different venues, when it came to data, we were quite lacking in having a 360-view of our customers. We were addressing their needs very well but we believed there were extra insights that we could build.
If you want to really compare it to pure retail e-commerce, their solutions are very mass, with big data and automation. Automation wouldn’t really fit for us, we’re looking for a truly differentiated approach. But where it can deepen our connections with our clients even more, that is the hallmark of LVMH service — excellent and effortless. That’s what our technology is for.
So how does that take shape in your data platform, as it develops?
We wanted to maintain the one-to-one interaction our customers expect from our maisons. It’s about weaving together data and AI that connects the digital and store experiences, all while being seamless and invisible. We want our customers to get even more dedicated attention without being intrusive.
So as I say, we need something more discreet than big data. We call it “quiet tech,” like our version of “quiet luxury.” We are not a tech company, but we also can’t just leverage the tech which is made available to us, either, because of the unique needs of the luxury world. That’s when we decided we needed to develop our own algorithms with Google, to identify our best customers, to identify the best product affinity for them, their taste and interests.
And then, because we are serving 75 different maisons, we also wanted to make this technology not so complex to deploy in all of our brands, while maintaining the autonomy of each maison.
One of the things that was appealing about Google Cloud was, we would have the ability to have this flexibility of implementing our own AI, our own data analysis, and the simplicity also to replicate that even across the smaller maison. A big part of this ease came from having BigQuery as a core component of the data platform.
So once this central team was up and running, what benefits were you seeing, and what challenges did you have to overcome?
This is the beauty of an organization like LVMH. Each brand is very distinct, with its own unique customers, though there is often overlap, too, at least across the groups, like fashion to spirits to leather goods or jewelry. If we pool our resources in the right way, we can do things, like this data platform, that none of our brands could take on on their own, at least not at the scale and effectiveness we have.
Take for example generative AI. We had built the core of our data platform in 2021, 2022, and that left us sitting at the ready when generative AI happened — we were already sitting in an organization which was very data aware. And so we were able to leverage that to really adopt generative AI. We’ve been exploring in Vertex AI, for example, the agents and applications available there, and how we can adapt them, like the data platform, to building custom agents that again meet our more unique needs.
What’s an example of some of the AI agents you’re working on?
Well, even when we step into the digital world, our motto is still to drive to the store as well — digital is not the end of the story. But we do need to adapt also the way we are interacting with our customers, to meet them where they are, which is so online these days.
We had built the core of our data platform in 2021, 2022, and that left us sitting at the ready when generative AI happened.
Crucially, you still have a client advisor, and that’s who we want to empower with data and AI. They have a list of clients to talk to, and this connection with data and AI helps them better to understand the client’s product affinities, and guide them to a better purchase or experience the client may not have known about. Also, as a client advisor, you may have someone who just landed in your client list but you don't know them well for many different reasons. Because we have a client history, through the data platform, we can provide a fuller picture and lifetime service.
And then we layer on an AI agent, and the advisor can chat with this data to find what they need even faster and be better informed, and it can even make new connections for them, new opportunities for additional items or experiences. And this leaves more time for the advisor to focus on the relationship with the customer, as well.
You mentioned that each maison and brand is building data and AI products on your platform in unique ways. What are some of those?
Our maisons have always brought a personal touch to the client relationships, so the hyper personalization that a lot of tech is delivering, everyone is really looking to go beyond that. The needs of Tiffany and Sephora, of Dom Perignon and Loewe — each of our maison are distinct. So we’re really building a platform that can serve all of their needs at once, providing connections where appropriate, but we always maintain a firewall between the brands and the data they’ve collected. Each maison is still in charge of the implementation and support of their own general IT ecosystem.
Now away from the customer data, there’s plenty of sharing between our maisons. In finance, HR, supply chain, areas like these, we’ve built benchmarks to help the different brands to compare each other. They can see how they’re performing on digital conversion, for instance, and then we can discuss that together and see how the maisons might benefit from the discoveries of one another. It’s playing a role in sales forecasting, supply chain forecasting, a lot of these areas where we can build efficient AI algorithms and look for opportunities.

We are trying to use data and recommendations in ways that are unique to us, like determining the best stores for certain pieces of very high-end jewelry, for example, which there may only be one of or a few of in the whole world. So we have to think carefully about which stores get these pieces, and then we want to use the client data to let them know, “Maybe next time you’re in Zurich, you may want to stop by the store and try this on.”
To some degree, it’s the same tactics as others, just used in ways that are ultimately unique to LVMH.
And then how about at the maison-level?
To take one big one, Tiffany’s has been one of the leaders in working with the client advisor agents we talked about, knowing the client, and proposing new pieces or offers, through the client advisor, it makes suggestions based on the data we have.
Are there more conventional ways that LVMH is deploying generative AI?
There are the more common gen-AI retail use cases, like crafting product descriptions for our websites or assisting with customer service inquiries. These are common across the industry, but in a way that is becoming the baseline for every brand, mass or luxury — you simply have to do it for competitive and financial reasons. We’re taking a “try it, deploy it” approach across the board, with lots of use cases, to see what works.
Our generative AI platform, some of it sounds simple, but it’s really doing amazing stuff. We have more than 40,000 users per month around the globe doing more than 1.5 million queries, and we know that it's providing a lot of efficiency, say for digesting documents or creating translations, it’s huge. Where AI can further deepen the excellence of personalization in our customer relationships, or the operational efficiency of our teams and collaborators, that is the hallmark of LVMH teams. That's what our technology is for.
Ultimately, we want all our technology to support the excellence of our approach to customers and the uniqueness of our products in a quiet, efficient and innovative manner. Like fine tailoring, we want our IT to be almost invisible, serving our products and brands to the delight of our customers. This has been and will continue to be the core of our partnership: serving our products, our collaborators, and our brands.