AI is not a cannibalistic infrastructure, it’s a creative medium: Vêtir CEO

Luxury and tech have long operated in different worlds—one built on emotion and exclusivity, the other on data and efficiency. But AI is forcing a reckoning on this frosty dynamic.

Vêtir, for instance, is an AI-powered personal styling app that is working to change what luxury looks like in the age of agentic commerce.

During Retail Brew’s recent virtual event, Vêtir Founder and CEO Kate Davidson Hudson, talked about the state of the industry, Vêtir’s AI playbook, and how AI can lift luxury fashion, enabling stylists to curate better.

This interview has been lightly edited for length and clarity.

Has the luxury and tech relationship changed, or has AI made the distance even wider?

Right now, with everything happening in AI, it’s not that that relationship has been bridged in terms of ideology; it’s just that the stakes are too high that neither can really sit on the sidelines anymore. And what’s really important for both cohorts to understand, especially the fashion world, is that AI is not really as it’s been marketed as this cannibalistic infrastructure. I think that the way that brands and designers—at least the ones that are more forward-thinking—are starting to look at AI, especially in an agentic form, as more of a creative medium.

If you were a designer and you’re designing a collection, you are taking into account the cultural context, who you’re designing for, how they’re wearing your clothes, what price points make sense, and now AI has really just empowered them with this kind of superpower that can deliver all of that data at speed.

What is Vêtir most focused on when it comes to agentic capabilities and that AI playbook?

We’re very focused on that post-purchase intelligence layer. Since the advent of e-commerce, there’s been such a push and such a focus on, “How do we optimize for intent? How do we take that consumer and drive them to the point of sale and drive them to convert and to transact?”

But if you look at the life cycle of a product and how you own the product, that’s such a small sliver of time in the life cycle of the product that you own, and there’s so much rich valuable data in what happens post-purchase. How does that consumer integrate that product into their daily repertoire? How does that consumer style a piece with other pieces in their closet? In what context, in what scenarios? Is it travel, is it work, is it an event?...Or did a client buy 10 pieces last year, but they only wore three, and why is that? All that information can provide an incredibly rich feedback loop, not only to brands and retailers, but also to stylists and to marketers, and that ultimately informs also our agentic AI models.

It’s something that historically brands and retailers have been completely blind to. So, a lot of the infrastructure that we’re building and optimizing is focused on that piece.

Most AI tools like ChatGPT and Claude are built for mass consumption, so how is Vêtir building a tech stack that actually serves the luxury customer?

We think of it less as a tech stack and more as three layers of intelligence that cross-pollinate and work together. The first one is obviously the fundamental piece of the digital closet. Every item that you have in your closet is extremely rich with data, and also that’s not a static element; that’s a completely evolving always-on kind of data feedback loop on who you are as an individual, how those pieces interact in your closet, and what your evolving style profile looks like.

The second piece is the human layer, which I think, candidly, is where so many of these new AI styling or shopping platforms get it wrong because they’re actively working to take the human out of the experience. Especially at the luxury end of the sector, it’s not about the product; it’s about the relationship with the product, and that’s a humanistic exercise.

To understand the emotional quotient, the forward-looking, instinctive emotionality of a piece, we have stylists who train our model because a lot of these LLMs are looking at a past data pool of information in order to derive recommendations. But fashion, by its nature, is forward looking, so it’s really that human input on making recommendations and distilling what’s happening in a fashion show that happens four months before that product lands on shelves.

So the [AI] agent, because it’s been trained on such a high-intent, luxury, rich data pool, is able to surface recommendations and outfits and cues to the stylist in terms of how to optimize their business.