The Edit is our new weekly show, where Social Editor Grace Robinson quizzes editor-in-chief Ben Hanson on some of the most significant fashion and technology stories from the past seven days.
This edition breaks from the usual format: Danit Peleg, Founder of 3D Printed Fashion Lab, joins us for the first half to explain how AI prompting now sits at the centre of her design and production process, and what still has to be true for 3D-printed garments to be adopted at scale.
Then Grace and Ben return to the usual setup to talk about Salesforce’s investment in Callimacus, and the case for pageless websites assembled in real time; SHEIN’s new prospectus, its thin margins, and whether the company is less unique than the industry assumes; and Daydream and THG Ingenuity both packaging their AI capability as a service.
This show is available in video format, embedded below, as well as audio, on Apple Podcasts and Spotify.
For deeper interviews with brand and technology leaders, our prestige interview show releases on Thursdays everywhere you find podcasts – and will be coming to YouTube soon.
Subscribers to our mailing list will receive notifications for both shows.
Grace Robinson: So welcome to The Edit, from The Interline — the show where we run a quickfire analysis on our pick of the most important fashion and beauty technology stories from the last seven days. I’m Grace, the Social Editor, and I’m joined by Ben, the Editor-in-Chief. And together, we have less than twenty-five minutes to give you our analysis on the stories that we think really matter.
Ben Hanson: Hi, Grace. I’m still travelling this week, as people will be able to see. We also have a bit of a break from the usual format this time. Normally it’s you quizzing just me. You are going to quiz me afterwards on a couple of news stories, but we have a separate guest today, who I’ll hand over to you to introduce — because she’s going to answer a couple of questions about her own work that I think have some broader implications for what people see in the news as well.
Danit Peleg — AI prompting inside a 3D-printing lab, and what mass scale still requires
Grace Robinson: Yes. So, like Ben mentioned, on with us today is Danit Peleg. You’re the Founder of 3D Printed Fashion Lab, and your journey with 3D-printed fashion has been pretty amazing. You started while you were still a student, and now you’ve been featured in Forbes and on the BBC. You’ve also grown a massive following on social media through sharing your 3D-printing process. But what we really wanted to talk about today was a recent talk you gave for Figma’s Config conference, where you shared how AI prompting is now part of your design process, and what you’re doing with 3D-printed garments. So, as a first question, maybe you can introduce yourself very briefly and also give a thirty-second summary of the talk you gave at Config.
Danit Peleg: Yes, absolutely. Pleasure to be here today. My name is Danit Peleg, and I’ve been doing 3D-printed fashion for the last decade. It started as a school project. I studied fashion design, and I was always curious about bringing different fashion technologies into my work so I could create more custom-made designs. So it was pretty natural for me to explore the world of 3D printing. It was fascinating to see that no one else was actually using these small desktop machines to print textiles, and so I had to really teach myself everything I know about this technology.
Fast forward, and I started public speaking, talking about my work, teaching at fashion schools, and working with major luxury fashion brands and sports brands to develop their ideas into 3D printing.
I’ve always been researching better materials — the materials aspect is something that is really important to me. My recent development is with two partners: we develop materials that we can 3D print that are made from algae, castor and hemp oils. So they’re fully biodegradable, with zero microplastics, they can come in any colour, and they’re very soft for 3D-printing textiles.
At my talk at Config, I’m showing how I use AI in my processes — in my design process, but also in the production process. Since I started working with my agents, it’s really empowered me on so many levels. Now I can actually make the whole process completely automated. A 3D printer, at the end of the day, is just robotics, so they work so well with the AI agents. At some point, I will be able to operate the machine by itself, using other robotics that will pick up the plates. I’m also showing that at my talk at Config. But mostly I’m showing how it can start from a concept and become an actual physical garment, using AI in my workflow.
Grace Robinson: It’s so amazing what you’re doing. Obviously it’s super innovative, and you’re really accelerating how we can produce 3D-printed garments. But what would have to be true in order to use this technology on a mass scale — to produce 3D-printed garments this way at mass scale, so they’re widely adopted and become garments that we wear day to day?
Danit Peleg: So we’re getting there. AI is empowering it and becoming so much smarter, literally day by day. My agents are becoming smarter. They know how to do the 3D modelling. They worked with me on software where I can take your idea and, without seeing you in person, make you a dress that fits you like a glove, and then generate a file that I can send you on WhatsApp so you can print it on the other side of the world — or I can send it to print in my lab.
Not a lot of things need to happen. I think when it comes to 3D-printed textiles, we do need to improve the hand of the fabric, so they will be even softer and will feel more and more like the fabrics we know today. Today it feels a bit like a power mesh fabric — if the audience are familiar with that kind of fabric, you can imagine how it feels. But I can print any structure and translate any idea. So really, there is no limitation but our imagination using this technology.
Grace Robinson: And how has AI accelerated what you’re doing? What’s the biggest thing it’s helped with?
Danit Peleg: Great question. If we need to choose one, one of the biggest benefits is that it knows the printer I’m working with so well that the code itself is of such high quality that I don’t have any mistakes in my prints. It’s generating my code — at the end of the day, the garment is just code. So by integrating it with AI, I can get better quality, faster results and less use of materials. Now, every time I print a garment, I print it to shape. I don’t cut anything. I just print whatever I need, to the exact size. And then it’s really easy for me to assemble the clothing together.
But the whole process starts with a prompt. The prompt I give to my software is a picture of a pattern. So let’s say I’m taking a floral print, and I put in the pattern making — the shape of the garment — which is a traditional fabric file. It’s called DXF. It’s just a normal file for pattern making. Today, almost nothing we wear is made with ruler and paper any more; we use software.
So I’m taking the files that this software can produce and putting them into my AI tool. And then the rest is done by the AI. It knows how to translate the floral print and the DXF file and merge them together. And then I just tell it which printer I’m going to use for this work. So the G-code is perfectly done.
Grace Robinson: Amazing. That’s all the questions from my side. But Ben, did you have any other questions for Danit?
Ben Hanson: No. It’s just been a fun thing to see. Danit and I met years and years ago in New York — I think it was a kind of innovation event. And you’ve been lodged in my head as a 3D-printing pioneer ever since. So the reason we invited you on was that it was fascinating to see you pop up at Figma Config, talking about just how embedded AI has become in the same kind of process. I think people think about 3D printing as living somewhere on its own, like a little island — experimental, and progressing.
But I think maybe people don’t realise quite how embedded it can become in the overall process, or the role that AI plays directly in it. So it’s been great having you join us for ten minutes of this show. Danit, very different format from usual. I intend at some point later in the year, or early next year, to get you on the one-to-one interview show and go a little bit deeper into these things. But this has given the audience a quick overview, and we really appreciate you joining.
Danit Peleg: Thank you so much for having me.
Callimacus — Salesforce backs the pageless web
Grace Robinson: We still have fifteen minutes of our usual recording time, so I do have a few news stories that I want to talk about. I’ll get into those now, and Ben, you can give us your answers. One thing that stood out to me this week was the announcement that Salesforce has invested in Callimacus — an AI platform backed by shareholders including Brunello Cucinelli, who I know you’ve spoken to, or about, on The Interline before.
What caught my eye about this story is that the company has this vision of pageless websites and applications, where AI agents understand each visitor’s intent and can assemble personalised digital experiences. This really got me thinking: do you think this is a new era of real-time, personalised ecommerce? And as customers’ expectations evolve around this, do you think big brands need to invest in AI to create these kinds of digital experiences for ecommerce in order to stay competitive and remain relevant?
Ben Hanson: So the relationship between this product and Brunello Cucinelli, the brand, is a slightly complicated one. I interviewed Francesco Bottigliero, who is jointly the head of humanistic technology at Brunello Cucinelli but also the CEO of a company called Solomei AI, which is headquartered in Solomeo — the town that is the headquarters of Brunello Cucinelli — and is part of the Cucinelli family holding company, the Foro delle Arti. So it’s a complicated one. It’s an instance of a brand developing an AI project that it deploys for itself.
The first place Callimacus is used is on Brunello Cucinelli’s corporate website, and on Brunello Cucinelli’s ecommerce website as well, as the default mode. So you can switch back to a traditional storefront, but they deploy this as the straightforward one. Now, I would encourage anybody who’s interested in this to listen to that conversation with Francesco. It’s about an hour long — it’s one of our deep interviews — but he gets into some of the philosophy behind this, how it relates to the experiences they want consumers to have, and what it means for a capitalist brand to deploy this sort of thing.
It’s a fascinating conversation, one of my favourite interviews, and I actually have Francesco joining me on stage in Paris at Première Vision in a couple of weeks, on 1 September, to talk about this. Now, the story specifically here this week is that not only have they deployed it successfully within Brunello Cucinelli’s own operations, I think it’s been clear to them for a while that it’s also a viable product — a viable platform to deploy across other brands as well — and that other companies are interested in it. So this kind of investment, and this wide recognition of it, was inevitable at some point.
Specifically what it is, again, I’m really summarising the interview here, but it is composable ecommerce storefronts, composable web pages. So traditionally, if you and I go to an ecommerce storefront or a marketplace — if you open amazon.co.uk and I open amazon.co.uk — even if we like the same things, we’re going to be shown different products upfront. That’s traditional A/B testing, where personalisation happens at the offer level: show this guy some of these things, show this lady some of these things, because that’s what they’re interested in and they’re more likely to convert. The Callimacus approach is that the page itself is actually built for you in real time, at runtime. So what you end up with is not just, okay, I’ll scroll through the same wireframe that you do but it’s going to have different stuff in it — the wireframe itself is going to look different.
And the reason this is so interesting to me is that people talk a lot about AI as the highest-intent and most personalised conversion channel. We’ve talked about this before on this show. Theoretically, if I’m having a conversation with a chatbot and it then recommends that I go and visit a brand or retailer storefront, that conversation is loaded with context about what I want. And if you’re able to pass that context over to the brand or the retailer, they should be able to show me something very personalised to what I want. The bet here is that they should be able to actually build the website — build the page — in real time.
And I think if you look at this from Salesforce’s point of view, or from any other retailer’s point of view, that’s a desirable thing to do, because you can put high-value inventory in front of people who have high intent, and the odds are that they will convert from there. I think there’s a much deeper conversation — it’s one I’ve had with Francesco, and one we’re going to have again in the future — about doing the same thing for software and applications as well. You sit down, we’re recording this on Riverside, but say you want to sit down and interact with a CRM, an ERP, a PLM, or what have you: having it actually just be a library of tools on the back end that then assembles itself in real time to what you need is fascinating, and I personally do think that represents the future of software in general.
SHEIN’s prospectus — a retailer like any other
Grace Robinson: So the next story is about SHEIN. I know SHEIN has been in the news a lot lately, and we’ve also previously spoken about SHEIN’s IPO. This week there’s been a lot of talk about SHEIN again, because the company issued a new prospectus, and analysts think it shows slowing growth and lower profitability than people might have expected. We tend to think about Shein as a completely different kind of business, and that’s why it’s been so disruptive and different to other brands and retailers out there. Do you think all this news and all this talk supports that theory — or do you think it’s maybe evidence that SHEIN is less unique than people might think?
Ben Hanson: So I’ve been on the record about the way that the EU, France and so on have approached SHEIN over the last year or so, saying I don’t think you can regulate away demand, and I don’t think you can regulate away a business model. I still stick to that. But obviously that attempt at regulation, and the desire to rein in SHEIN that you see across Europe, in America and in the UK — where you’re trying to close things like de minimis shipping loopholes — is all based on the understanding that SHEIN is something different. It is a huge-scale, massive-value overseas disruptor to domestic business models that regulators and governments think have an inherent disadvantage when they’re held to a single country. They have a disadvantage in terms of duty. They have a disadvantage in the sense that they are reliant on the same kind of overseas supply chain as everything else.
So the assumption has been that the playing field is uneven, and that SHEIN has an advantage. What we’re seeing now — you and I talked previously about SHEIN’s IPO target being about half of what people expected it to be a couple of years ago. The new analysis that’s out there is based on a prospectus that I believe was published at the weekend, last weekend, by the time people listen to this. And what that prospectus shows is that for all of its scale, for all of its speed, for all of its touted disruptive potential, it makes very thin margins, and it makes a lot of products — it floods the zone, as it were, in terms of inventory and everything else.
In other words, SHEIN is a brand and a retailer in the same way as any other. And I think what this represents in a broader sense is that there’s been this idea that these digital overseas disruptors feel more like tech companies, so they should be valued more like tech companies, and they should be feared more like tech companies. The more we see behind the curtain, I think the more we realise that that tech advantage only gets you so far, and that you are still in the same business of fashion that everybody else is in, rather than a fundamentally different one.
So for companies that have been disrupted by SHEIN, companies that fear it — we talked last week, I think, about innovation and advancement in the value space, and how nice that is to see. And I would take this now as further evidence that extreme value, affordable, high-volume fashion is up for grabs. If you want to play in that space, don’t consider it as cornered by SHEIN and Temu. There’s still plenty to go after here, because they are potentially not as scary a company as people have made them out to be.
Daydream and THG Ingenuity — AI capability sold as a service
Grace Robinson: So the final story I want to talk about is about two announcements we’ve seen this week, and they’re actually in a similar space to what we just spoke about with the Callimacus story. The first is that Daydream, who I know you’ve interviewed before, have announced that they’re going to be offering their platform directly to brands and retailers to deploy on their own ecosystems — so the model they’ve created, but delivered as a service rather than a partnership.
And there’s also been an announcement that THG Ingenuity, who you’ve also covered when you visited THG Studios a few months ago, are offering what they call their AI Stylist, which is a virtual try-on, in partnership with Google, as an app on Google’s Cloud Marketplace, so anyone can use it. So I wanted to get your take on all of this — do you think there’s a pattern here?
Ben Hanson: So I do think there’s a pattern here, and part of it is that there are in fact two. One is the pattern those companies would like to talk about, and one is the pattern I don’t think they would. So when I interviewed Daydream, I interviewed the CTO, Maria Belousova — I think it was towards the end of 2025 or early 2026. A very smart woman, who has put a lot of thought into how people interact with AI and how that then feeds into product discovery and recommendations. I hate to just keep pointing people to other interviews, but I would encourage people to go and listen to that one, because there’s a particularly pertinent thing that Maria said: that there’s a wide variety of cultural variables and other things that go into the consumer vocabulary people use to express what they want from fashion.
So people don’t just go in and say, I am looking for a dress with this neckline and this material, it needs this construction, it needs this wear, that kind of thing. What they will say is, I want it to look like this thing I saw in this film, I want to wear it in this particular use case, and so on.
Now, historically, you would have to do that through Daydream directly, as an application. So you would have to download Daydream, you would have to have your interaction with it. Their catalogue is populated, I believe, with brands that they have previously partnered with, and at the end of the conversation they will say, here you go, let’s present you with these options. You can then refine them more visually, and so on. This turn now is them saying, well, let’s take that same technology layer and that same understanding of interaction and interface, and let’s make it available as a product that brands and retailers can buy and deploy on their own, rather than having to have their consumers go through Daydream.
Now, it’s a good business turn for them. It gives them a new market. It’s interesting for retailers that have not either wanted to, or been able to, build this kind of thing for themselves. It’s also, to some extent, an admission that that sort of platform has maybe not secured as much of the user base as it could have done on its own. If you’re creating this kind of business, your initial play is: how much of this can I own? How many people are going to come directly to my app? How many people come directly to my platform? How can I monetise handing that off to the brands afterwards?
If you do that successfully, if you get sufficient amounts of it, you don’t then need to sell your platform any other way. Selling your platform another way is a great additional revenue stream, but it’s also a testament to the fact that I think AI interactions like this are not going to become balkanised into individual apps the way that people maybe thought they would be.
The THG Ingenuity — that’s a hard name to pronounce — play is a little bit different, but it’s fundamentally the same thing. So THG stands for The Hut Group. They’re a massive multi-brand company here in the UK, but they also offer fulfilment as a service, warehousing as a service, and a content studio as well. It was the content studio that I visited for the Topshop AI catwalk thing a couple of months ago. There’s a photo somewhere on that article of me actually trying out this service on a kiosk that was there.
They make it available through Google Cloud Marketplace for basically the same reason. They can do this as a service for brands, and their initial intake is companies that want to come and work with them and have them provide this directly. There’s a limit to how far that scales, and I think if you are a provider like that, putting it on a marketplace means people can then just get it through the Google Cloud platform, deploy it across their own infrastructure, and so on.
Again, it’s another smart way to say that AI is able to change one or more elements of the way people interface with brands and products. It may not be compelling enough in its own right for people to go and interact with different applications, and to do their shopping in multiple different little pockets. But if we’re able collectively to just uplift the whole infrastructure and put this in as many places as possible, that’s probably likely to lead to more change long term.
