Episode Transcript
Kyle James (00:01.07)
Hey, welcome to the AI Chronicles podcast. I'm your host, Kyle James. And today we'll be discussing how an IT and fashion analytics company called Prevore is using AI inside of their own business. And we'll share the exact steps that you can take in order to implement AI for yourself. Now, before we talk about that, listen closely. Are you looking to implement AI inside of your own company? Or maybe just struggling to get your AI to stop hallucinating? Speak to GPT Trainer.
GPT Trainer literally builds out and manages your AI for you, eliminating hallucinations for good. Go to gpt-trainer.com. I promise you, it'll be the biggest time-saving decision that you've made all year. Trying to set up AI on your own is like trying to build a house from scratch. Sure, you could do it, but the time of frustration is gonna take you to get it finished. It may not be worth it. It's a thousand times faster and safer to hire professionals.
Schedule a consultation today. again, that's gpt-trainer.com. Say hi with me on the show, Courtney Cost, who is the founder of Prevore, an innovative AI-driven planning tool launched in 2024 that's transforming the fashion industry. Courtney has over a decade of experience across fashion, retail, and luxury merchandising. Her company, Prevore, leverages computer vision and machine learning.
to help merchandisers analyze product performance, optimize assortments, and forecast trends by extracting details from sales data. Backed by Atla ML, Courtney's mission is to replace intuition and fragmented data with intelligent insights that reduce waste and boost profitability for luxury brands worldwide. Hey, Courtney, welcome to the show. How are you doing?
Courtney (01:50.337)
I'm great. How are you, Kyle?
Kyle James (01:51.983)
I'm doing pretty good. for asking. So give us some context here. Tell us a little bit more about Prevore, like how the company got founded and what exactly are y'all doing over there at your team?
Courtney (02:01.889)
Absolutely. So the thing about Prevoir is we're not a company that's set out to be an AI company. And I'll give you a background of why I say that. So back five, six years ago, I was in grad school. I have a background in marketing and fashion, and I identified the use case that is what Prevoir was built around. So in my academic research, I learned that fashion businesses and fashion brands are swimming in mounds of data.
but they don't have really great tools to analyze their data. There's a lot of inventory planning tools out there, but they're not fashion specific. They tend to wrap up retail as a whole. And so there's no real solution out there that's very fashion focused. So, you know, through my academic research, I talked to many, many brands from fast fashion brands to very high end brands like Chanel or Louis Vuitton.
and almost all of them use Excel spreadsheets to analyze their data. So as I said, that was about six, seven years ago. And so the last few years was when I really realized that AI could help solve this problem. And so I became fascinated with computer vision. I learned about a few other companies, French companies that were using computer vision to analyze fashion trends on social media. So they would...
use computer vision algorithms that they had built and they would analyze images typically on social media, so Instagram, Pinterest, and they were forecasting trends in that way. I became very fascinated with how a brand could use computer vision applied to their internal data. So what Prevoire does is we import the sales data for a fashion brand, which
typically lives in a system like Shopify or Demandware, and we augment that data with computer vision attributes. take for example, the shirt that you're wearing, we would be able to pull the fabric, whether or not it has buttons, the sleeve length, the cut, the length. And when we combine those natural language attributes with the sales data, we're able to augment the sales data and give brands a better idea of which attributes are performing best.
Courtney (04:22.969)
And this is really built around a very manual process that brands already do. Like if you're a merchandiser for a brand, you're looking at your sales data and then you were just physically looking at your products and you're analyzing, you know, with your own eyes, why you think one product has performed better than another. And as we all know, whether you have a lot of fashion background or you don't, that fashion is an inherently visual...
industry. So really what we're doing is we're creating that data set as opposed to merchandisers just looking at clothes and trying to decide for themselves why something has done well or not.
Kyle James (05:03.084)
Yeah, for sure. And so I love the computer, computer vision that's like being integrated into this and even on the AI front, like what's kind of like the common cause you mentioned a little bit, like, you know, these, companies have so much of this data and they're just eyeballing it. then like, walk me through like, kind like the transition when you made that shift where you're like, okay, we are going to start implementing computer vision. We're going start implementing AI. Like when that shift happened, like what were some of the biggest like elements or key factors that made you go, wow, like this is completely different.
And it's game-changing the way we're doing business.
Courtney (05:36.857)
Um, well, first of all, computer vision has gotten so much better even in the last year. But like when we first started experimenting, um, with, you know, training data, just images we pulled off the internet, um, even just with the improvements of chat GPT four, oh, it made it so that something that we thought would take six months and a whole bunch of training data could be done, you know, within a few days. So computer vision has improved so, so much, and that has made our business.
not just easier to do, but way less expensive, way less labor intensive. But for fashion brands, they're a bit laggards in terms of adopting new technology. So we've had to put a lot into the user interface and not just having this data, but being able to show in a very user friendly way, like how they can apply this. So we've spent probably nine months just on the user interface.
And that has been a really big game changer for us. So it's not just the data. It's like, how do we show that data in the best way possible so that brands can just get up and running very, very quickly and don't have to have a really long onboarding or training to be able to integrate this data into their sales data. So that's been a really big game changer as well.
Kyle James (06:55.534)
Yeah, for sure. do you think that like, I mean, obviously you've made this shift right on the AI front and with overcoming some of these challenges, if you didn't implement the computer vision or the AI in the first place, like, do you think that even you'd be where you're at today? And like, do you think you'd still see those same types of like results that you're seeing now with some of your clients?
Courtney (07:17.921)
No, I don't think so. mean, there's other solutions that are, you know, as I said, data analytics platforms for retail verticals, you know, from fashion to beauty products. And there really wouldn't be much of a differentiator for us. But surprisingly, there's very few companies that are implementing computer vision in this way. There's a lot of other use cases for computer vision in fashion, like I mentioned, trend forecasting with external data sets. But yeah, very few companies are applying it in this way.
And when we're able to do a demo for a brand, they're really, really impressed because it's just a very natural thing that they're already doing. And now they have, you know, an assistant to help them do that.
Kyle James (07:57.166)
Yeah, that's very cool. So why don't we kind of through step by step here, like if say, a, know, a popular retail brands, like, Hey, actually I want to, I want to start working with Prevore here. Like here's all the data. Like what's, what does it look like from start to finish? Like when you, when you onboard and then kind of like what's happening with their data, what it's doing on the computer vision side, and then kind of the final outlook of kind of spits out like, Hey, here's the final output. This is the, the revelation that you're getting from all this data.
Courtney (07:59.673)
you
Courtney (08:25.849)
Yeah, absolutely. So first of all, we can only work with brands right now that are on Shopify. Almost half of fashion brands are on Shopify. Looking into 2026, we will be able to accommodate other sales platforms. So if you're on Shopify and you have a fashion brand, then you can use Prevore. We're really proud of the fact that it takes us about 30 minutes to onboard a new brand.
You're not going to have all of your data populated in 30 minutes, but the integration is very, very quick. And then, you know, typically within a few days, all the attributes have been pulled from your products, depending on how many products you have, of course. And then when you log in, we have a few different dashboards right now. Another thing I should add is like we're combining, you know, the sales metrics with the attributes. So you're able to see what are your top performing colors.
based on a date range, based on a category, based on a specific brand, if you're a multi-brand company. But we also give contacts to that data. So let's say black performs best for pants in wintertime. But we also show you how much did you lean into that? How much did you allocate your inventory to that color? Should you have allocated more or less? That's one of the things that we've really learned.
testing with the brands that we have and our pilot brands is that it's not enough to just show the data of, how did a certain print of dresses sell for you, you need context for that. Otherwise, the information isn't really all that useful and we provide that context. So you're able to see all of your products and you're able to see those metrics by product, but you're also able to make comparisons and also time comparisons are very, very important. Like maybe you want to see how
a certain fabric did for dresses in spring of 2025, but you need to be able to compare it to spring of 2024. That's a very common comparison that fashion brands are constantly doing. They're constantly looking backwards up to about a year to see where they're doing in terms of their benchmarks.
Kyle James (10:35.822)
Yeah, for sure. what, since you've been working with different clients like that are obviously integrated within Shopify and getting feedback within the whole process, what kind of results have you been seeing so far that's maybe something that's worth highlighting to some of those, especially those out there who are like, I've got a similar business here. This could be something that's a big good fit. What are some of those key highlights you would mention?
Courtney (10:38.669)
Thank you.
Courtney (10:58.787)
Well, we're really targeting brands right now that have a pretty good amount of revenue, like 20 million plus, but they're still running pretty lean. Like the fashion industry, if you Google fashion news today, the fashion industry is struggling right now. Revenues are down except for a few very specific hero type brands.
And so over the last few years, especially since COVID, brands are really feeling the pinch in terms of leaning their teams out. And I should, you know, cave it as well. Like we're in beta. We're not completely finished building Prevoire. So we have, you know, pilot brands using Prevoire right now. But the biggest piece of feedback we've gotten so far is that they can keep their teams leaner.
There's a lot of KPIs we love to hit. We'd love to be able to prove that we can reduce the amount of dead stock you have at the end of the season. Really be able to prove that we're saving you a lot of time. We're giving you better analytics than you're getting in any other platform or even within the Shopify platform. But so far the early feedback is that they can keep their teams lean. And that's even why some brands have been onboarded on to Provoire exactly because they
They want to keep their team small. don't want to hire expensive consultants or they don't want to license really, really expensive software that needs a consultant because a lot of them do.
Kyle James (12:26.018)
Yeah, for sure, for sure. So you talked a little bit like it's kind of like in the beta mode, but you're building out this app. I guess what are some of those like upcoming AI initiatives that Pivora is planning for and like, where do you see it playing, especially in the fashion industry, playing the biggest role in your operations next?
Courtney (12:42.905)
We're experimenting a lot with agentic AI. We already have a beta data analysis agent. So functions very similar to chat GPT. You can ask a question like what were my top performing pants last week and it'll give you an answer. So right now we're really experimenting with our users to understand like what kinds of questions they're asking and what kinds of answers they expect.
I think that's probably the biggest room of improvement there. But so far we've found that it's not giving wrong answers, it's not hallucinating or anything like that. So that's on the horizon to really perfect that and make that really, really usable, not just in this test mode. We are also working on using AI to generate reports. That's a really big function for most fashion brands is they're doing reporting.
weekly. It's a very time intensive process. And a lot of them are doing manual reports are building in PowerPoint, they're building them just in Word. So we want to be able to automate those and make them just customizable enough, but not too customizable for them. We're working on range building. So applying AI to be able to take all this data that you know, we've spent the last year building out like we have the metrics, we have that
augmented data set with computer vision and now how can we use that to help a brand decide how much should you buy of something, what kinds of attributes should you lean into, whether it's colors or fabrics or cuts for the products that you're either designing or buying. So that's the ultimate goal is really being able to help fashion brands plan for the future. And one thing that a lot of people don't know about fashion because there's a lot of
information out there about trends, what's trending, what's not trending on social media and what have you is that most of what fashion brands plan for is from their own internal data. So external trends do play a part in some respects, but not as much as you would think. For the most part, they're iterating on past success and that's typically what drives their future success. The last thing that we wanted to use AI to apply
Courtney (14:59.541)
is to metrics. Like for example, I've heard time and time again with brands that use Shopify is that the Shopify metrics are incomplete and they leave a lot to be desired. They're typically pulling their data out of Shopify and doing their own metrics calculations within Excel. So one example is we want to be able to use AI to really accurately forecast quantities for a brand. Like for example, there's
another inventory planning software that will tell you to buy something in a certain amount just because you've sold a lot of it in the last two weeks, whereas that's not taking into account seasonality, holidays, like if your brand has a real spike around Black Friday or Christmas, not all brands do. It's not taking into account all these contextual factors into forecasting and we think that there's...
Kyle James (15:39.48)
Mm.
Courtney (15:55.243)
a place for AI to be able to do that.
Kyle James (15:58.03)
Yeah, it's very cool. like, um, one thing you said, like with the internal data side, like, think there's a lot of companies out there who are like, then I feel like there's not even just in fashion, but just they're like, Hey, we've got to figure out what's going on externally. And I love that the queue of like, Hey, actually we need to figure out internally first. Cause that's like, that's what's driving the next steps more than the external side of things. Obviously keep in mind what's happening out there, but the key thing is, let's look at our internal data first and see what, where's the, where's the first step that we can take.
Courtney (16:01.273)
Thank you.
Kyle James (16:27.394)
That's going to make the biggest change for us. I love that perspective on that. And thanks for sharing that, And as we start wrapping up this conversation, where would you recommend people go to? Maybe learn a little bit more about you, and then maybe a little bit more about Prevore that you'd recommend.
Courtney (16:41.625)
Well, our website is purvoir.ai, but I think more information is available on our company LinkedIn page as well as my LinkedIn. We have an Instagram as well. I recommend going to LinkedIn. We do a lot of writing. We share a lot of updates on what we're working on, what's new with Purvoir. So the link to our LinkedIn is on our website as well, purvoir.ai.
Kyle James (17:05.996)
Awesome. Love it. you, Courtney. great having you on the show today. I really appreciate your perspective in the fashion industry. It's definitely unique and it's really cool what you guys are doing over it on the computer vision side and even on the AI side. yeah, for sure. And thanks everybody for listening. Remember, if you're looking to implement AI into your business today, please don't try and do it yourself. The time and stress that AI could cause, it may not be worth it. Schedule a call with GPD Trainer and let them build out and manage your AI for you.
Courtney (17:17.952)
Thank you so much for having me.
Kyle James (17:33.558)
Once again, that's GPT-trainer.com. Signing off, signing up for now. Have a great rest of your day, everybody. Looking forward to seeing everyone on the next episode of AI Chronicles.
