Latest / The Interline Podcast / The High Bar For Disrupting Knitting

Transcript

The High Bar For Disrupting Knitting · The Interline Podcast

• speaker-0: Hey, welcome back to the Interline Podcast. Now, I don't want to trivialize anyone's work here, but there are parts of fashion, entire categories, in fact, that are just inherently less complicated than others. Take a block colour five-panel cotton monomaterial t-shirt. It's a relatively easy thing to grasp, in the sense that anyone could understand the design, the construction, the material, even if the sewing operations required to put it all together aren't exactly common knowledge. Then on the other end of the spectrum, you have knitting. Which is probably the closest analog our industry has to software engineering and execution at the product level, in the sense that designers create binary pixel grids that specialized software translates into machine readable code, a knit program, that then drives industrial machinery. None of that is instinctive or easy to grasp in the way that cutting and sewing are. Now, I'm nothing close to being a knit expert. And I say that with some confidence because knit experts are, like pattern makers, one of fashion's deepest. Drafts of scarce specialized skills, and both the floor and the ceiling for talent are incredibly high. You don't casually walk into the world of knit and just figure it out, is what I'm saying. All of which is why knitting, as a discipline and as an industrialized production process, has become so heavily concentrated and consolidated, to the point where a small number of hardware and software vendors and a narrow pool of specialist producers exert a lot of control over the market. If you want to make knitware or products with a knitted component to them, at anything approaching industrial scale, there's a very narrowly scoped way to do it. One that's deeply technical and one that's steeped in a lot of history. That all sets a pretty high bar for disruption. Not in the usual digital disruptures dilemma sense, where someone is coming from outside a sector, armed with technology, ready to tell an analogue world how to do things differently, mind you, not that knitting, as I've said, is It's digital already, in a pretty meaningful sense, more in the fact that it's all so entrenched, and from the outside at least, it's so arcane and complex that upending any part of it, let alone all of it, feels like an impossible order. But that's an order that today's guest, Garrett Gerson, CEO of Variant 3D, has taken on. His company's vision is, and I quote, to create a platform that allows anyone to knit anything of any complexity with any yarn using any machine anywhere in the world. So I wanted to bring him on to do two things. To update and challenge my own understanding of knitting and where it crosses over with some of our most popular topics of the last few years, digital product creation and AI. And also to grill him a bit on what he thinks it looks like commercially to do something that big, to push a completely different model into a category and a process that's as embedded as they come. So let's see what he had to say. Okay, Garrett Gerson, welcome to the Intelligent Podcast. speaker-1: Ben, thanks for having me. Excited to be here with you. speaker-0: Very much pleasure on this side of the table. Now, we like to start the show by getting an idea of what the person on the other side of the interview desk actually does day in and day out. In your case, you're the CEO of a software company, which is a fairly typical profile for these shows, not not super unique to be talking to somebody who runs a software firm. You're also someone who has a bank of their own knitting machines on a ranch in Malibu, which is fairly atypical. Don't think we've had somebody on who we don't think we've had somebody on with exactly that profile before. Walk me through how you spend your time. Like it can be day by day or week by week. Give me an idea of what your time looks like. speaker-1: Spend you know, a lot of my time is ⁓ working with customers, ⁓ sourcing new deal feed. A lot of what I do is ⁓ you know, really working with ⁓ current customers and then, you know, just speaking with people, ⁓ educating them what we do. ⁓ hosting ⁓ guests here ⁓ on property. ⁓ and you know, sales is the end of the day, I think is you know, a great CEO has to be a great salesperson. ⁓ you know, we do a lot of fundraising as well. Can't say that's my favorite thing to do. ⁓ but it is it's something I've may become, well, I guess the things you don't like are the things you should lean into, is what my grandpa used to say. ⁓ those are the things you should really strengthen. So I've really leaned into that and learned how to, you know, really fundraise ⁓ so we can we could grow the company. ⁓ road mapping is a big thing as well, right? ⁓ and then I've got an incredible co founder, ⁓ Will. And maybe at some point it'd be great to have him on too, because ⁓ between the two of us we really have taken an idea or a dream that everyone thought was well, quite frankly, the whole industry thought was impossible. ⁓ and made it a reality. And and we're, you know, it is a it is a long road, ⁓ but we're we're not adverse or I've never been adverse to ⁓ doing things that people think are crazy because I always think there's a big opportunity if you can break through on the other side of that. But day to day is, you know, I I you know I really am not a I I don't like to micromanage. I love we have an amazing team of ⁓ of I mean some pro some of the best engineers, both software and knit engineers and development ⁓ professionals, ⁓ within the company. So it's really just let's set the North Star, let's do our quarterly goals, ⁓ let's do our daily stand ups and and then just try to tap the wheel and really I'm here to run interference. But ultimately my responsibility is to bring in customers ⁓ and then meet with our meet with our investors and ⁓ and you know and speaker-0: Yeah. Good. Well, I mean, I I always like talking to readers. ⁓ I we are bootstrapped rather than funding backed. I think I would be terrible at talking to investors. ⁓ so yeah, do maybe that's one of those things I'll le lean into over time 'cause I don't like it. you mentioned you so you know, I mentioned the ranch, you said on property. Just for the aud for anyone in the audience who's not familiar with variant, walk me through what on property means. speaker-1: So my great grandparents started Calamigos ranch ⁓ back in nineteen thirty seven. It's a four hundred acre ranch, used to be a working ranch. Now we're a big hospitality company. That's how I actually got introduced to the apparel and garment industry. So about twenty two years ago when I left college, ⁓ I took over our family business, which it was super rustic, but I ended up building a ⁓ a vertically integrated apparel manufacturing for uniforms and textile company. ⁓ in downtown LA said about ten thousand square feet. And that's kind of kind of the origination of how I got here, why it's here in Malibu. ⁓ the other piece of that is sustainability ⁓ and kind of closing that loop and localization, like bringing manufacturing back more locally and ⁓ being more responsive has always been something I've been passionate about. You know, there's nothing, I think nothing speaks volumes to coming to a 400 acre ranch in the middle of Malibu and the Santa Monica Mountains and seeing a probably one of the most advanced textile knitting facilities in the world with some of the best engineers where we are run on Starlink, ⁓ working to take the whole it's actually it's a retrofitted horse bar, ⁓ to take it off ⁓ to take it off grid to show that industrial manufacturing could literally be stood up anywhere in the world. So like you could do it here, you could do it in an industrial park, you could do it on a farm in the Coxwalds or, you know, you can you name the place, you could you could make it happen. speaker-0: Yeah. And we're going to get into the domestic production side of things, I think, a little bit as we as we get on. The other thing I always try and do up front with you shows is to pin down a definition. Pin down something about the person's expertise that helps to ground the rest of what we then talk about. And the term I want you to define is knit to shape. But it actually runs ⁓ basically completely counter to the traditional method of garment construction, which is based on cutting pieces and sewing them together. Tell me what it means to you, that definition, and tell me how that meaning has influenced what you're trying to build. speaker-1: Well, I I mean, you know, it's interesting is if if you not from the knitting space, knit to shape sounds kind of weird, right? It's like, well, what do you mean by exactly what you're asking? Is ⁓ you know, what does that mean? I mean, d I got a lot of heat for this for many, many years, but I really do think this is the closest thing to 3D printing textiles ⁓ that we're gonna get to using these types of machines. It is it is additive manufacturing at its core. ⁓ you know, there's you know, what's so compelling about why I was why I was drawn to this. I didn't actually realize how complicated it was to be honest then, like eight years ago when I started this, the company and looked at the problem, I didn't realize the complexity of what I was trying to, you know, what I was trying to do. I thought, well, because I came from the cut-and-sew world. So today, when you design, whether you use clo or, you know, ⁓ any of these other kind of patterning softwares, you design in flat panels, you take ⁓ rolled goods, you put a pattern on top, you cut it out, there's quite a bit of waste. You then take a 2D, effectively a two D shape and wrap it around a three D form. You know, ⁓ knit to shape or three D knit. ⁓ or what I like to c what I like to actually call this effectively three D printing textiles using flatbed knitting machines, is ⁓ you know, it's it's really additive manufacturing. You're taking yard and we're so we're using computational design tools that we built internally, ⁓ to take a a three D any type of three D file. So right now we're working footwear, apparel, you know, a bunch of different and we'll talk probably talk about this later on too, but A bunch of different other product categories, but very complex geometries. We're able to pull that into our software. And what would take, you know, ⁓ sometimes weeks or months, or and in in cases where we're doing with products now, almost not possible ⁓ with, you know, kind of the the traditional CAD tools out there. You have to design pixel by pixel on a flat, you effectively wrap the surface, you tape it up, you wrap it, you then cut it out, flatten it, you measure all those pieces, and you take a two, a three D shake, flatten it two D into a 2D pattern and then you program it in 2D. We're going from 3D to 3D to the final product. So we're eliminating a lot of steps. ⁓ but knit to shape is is at its core, additive manufacturing in textiles because you're taking yarn, doesn't matter what kind, ⁓ you're ⁓ spooling it up on or threading it up on the machine, and you're knitting those pieces to the final shape of the product. And a hundred percent of that yarn ⁓ in that material is actually used. Just to you know, to finish ⁓ the covering of whatever you're you're creating. speaker-0: Yep. That's a that's a really good definition. I think it helps to cast it in contrast to the n traditional approach. Now to lean a bit further into the platform and software side of things, which you just alluded to. Earlier this year you launched Loop, which I think your co-founder described as a and I'm gonna use a direct quote here, ⁓ a platform for knitting anything of any complexity with any yarn using any machine anywhere in the world. That's a very tall order, right? Because knitting knitting, as you said, is a complex discipline. I wouldn't I would not class myself as an expert. ⁓ I come to this as an outsider. I know enough to know how hard it is, ⁓ as you as you just alluded to. It's also a sector where the hardware side of things is dominated by a few major suppliers and where a lot of some of the same suppliers are also working to own the software, both solo and in partnership with some of the leading 3D vendors you talked about. It's it's very complex and very consolidated, I guess, is the simplest way of putting it. ⁓ is that an accurate read of the knitting ecosystem? Am I missing something? And if so, how do you propose to upend basically all of that consolidation at once? speaker-1: So yes, so Ben, yes and no, right? ⁓ I think ⁓ so machine manufacturers, you know, typically, and I think this is across the board ⁓ in all industries, aren't all you're either you're an a you're amazing at one thing, but not both, right? You make the machine, but the software is a derivative to sell more machines. So it's it's about and and the software they've that is available now, it's very complex, takes years to master and learn. There's ⁓ you know, there's a lot of nuance to it. I mean, I know I taught myself it took me two years with a mentor, ⁓ to learn how to program. And I knew I realized that was the glass ceiling. ⁓ it's not so much upending the industry, but it's creating it's you know, effectively what I like to say is it's we'd like to democratize the space, create accessibility to this amazing method of manufacturing, which right now is a is it's a very minority subset of how we make products owned by a few very big brands. But if you know, and and this is kind of where my passion for sustainability comes in play. Is that, you know, if if we can get the majority of people manufacturing d designing manufacturing and thinking how to make product like this, then we can drive down waste. ⁓ so I mean that is always kind of been my core motivation in a lot of the things that we do. ⁓ you know, just to create accessibility democratization. But ⁓ but yeah, there there is a consolidation, right? There's well, I mean, right now we just lost ⁓ stoll, right? So stole is being, I think, acquired by another machine manufacturing company. Unsure, unsure of unsure of I have some thoughts of who, but unsure of yet where it'll end up. But now there's ⁓ Shima, which is a Japanese manufacturer of flatbed knitting machines. ⁓ there's Steiger, which is bought by She Xing, which is a Chinese manufacturer. And then you have ⁓ Santoni Lenati, right? The Italian company. ⁓ stole will will will probably fall into one of those those categories. So it's further consolidating that the industry. But all you know, no one's I well, I'm I don't know. Maybe no one's been crazy enough to try to to solve for this problem. ⁓ when I when I first ventured out onto this, ⁓ when I first ventured out to work on this and I met Will, you d I I can't tell you how many people Ben told me it's impossible, you're you're crazy and ⁓ you'll you'll never figure it out. It's too hard. There's too many there are too many variables to it. And that is totally true. And I ⁓ but if you boil it down Maybe this is where I, you know, as ⁓ you know, one of my the attributes I I like to lean into is I can take a very complex problem and then try to break it down. I also have an amazing co-founder that's worked hand in hand with me to do this, but we broke it down into steps and you just layer it like a cake, right? If you try to boil the whole thing, try to take on all of it, it's like it's like eating an elephant. You gotta start at one end and you work your way to the other end. But you can't do the whole thing at once. It's just about possible. So we started breaking it down into little pieces and then and you know, mapping it out and saying, okay, once we do this, it unlocks this. And then it unlocks this. ⁓ and it what it's done is it's actually opened up a whole new design methodology on how people will design in the future, how you will think about design. And at the ⁓ and in all honesty, this was actually a tool to allow me to build something quicker, quicker, faster, and bring it to market, because we were going to launch our own brand once we realized the power of what we built. It was like, ⁓ we should really pivot this we should open this up and give this to everybody. speaker-0: Fine. Well, let's talk a little bit about that democratization side of things, because that's something I wanna pick up on the the platform side especially. So I've seen I've seen Loop referred to, and I forget whether this is from from your your materials or from outside as as Figma. So the the famous ⁓ web design and interface design platform for Figma for textile design. Because Loop is web-based and because it leans into the WYSIWYG, what you see is what you get acronym pretty hard. Now, anyone uses Figma or any Had a comparable web native design tool or workspace, they will know that the value of those comes from speed, alignment, and accessibility. I'll tell you what I mean. Like it it's faster to design interfaces in a web-based environment with components and standardized tools. It's quicker than doing it offline desktop software that's that's non-y can't work in components the same way. It's better to collaborate on those kinds of things in an environment that everybody has access to, because they don't need specialized desktop software. And as a result of both those things, those tools have exploded the user base for design in general. I think it's important to note that most of the people who you who interact with Figma are probably not designers, or they are probably not historically designers. They're people who are looking at work that's been done on that basis, or the people who have the capacity or the taste to become somebody who works in design. Same same is true for Canva, right? But they are not people who would historically have used bespoke software for it. And if we le if we think about that from the point of view of Loop, if you're somebody who uses, uses or used the specialized software that you talked about, takes years to learn, fine. Like something like this makes it faster and more collaborative. If you're somebody who's never used those things, but you're somebody who's interested in knitware or you work in that sort of space, this is this kind of approach has the potential to put more power in your hands. I specifically saw an announcement from your side earlier this spring where the the variant team was able to design and knit something like 10 different iterations in the space of 24 hours, going from CAD files that were created in a browser to directly driving the production machinery, the 3D to 3D you talked about. So that sounds like you've got the speed thing in your sites. What about the other pieces? So the Figma for textile design needs to be collaborative in the sense that you're not just it's not just for a narrow audience. It needs to be for everyone. And I don't think I'm spoiling anything by saying that computational knit design traditionally isn't for everyone. So yeah what do you think it actually means or looks like to have a Figma for textile design? speaker-1: Well, I mean, I think the biggest thing is about I mean there's a couple there's I mean you hit all three, right? It's ⁓ speed, ⁓ accessibility, and then collaboration. And I think the big the the best thing about Figma, or my one of my favorite things about Figma is it ⁓ it creates a real real real real time space where everyone can come into it, come in, see, make comments, collaborate, and it creates a lot of ⁓ it creates a lot of visibility on a project. You don't have that right now when you design a creative product. ⁓ the biggest issue ⁓ for designers. So where we're entering in as our ⁓ you know as as the user we're focusing on right now is the designer. ⁓ designers typically need a knit engineer to create whatever sketch or ⁓ 3D model file that they've sent to this knit engineer. They need to wait then weeks to get an abstraction of what that's going to look like. Typically it's not it's a you know it's not what they want. So then it becomes a negotiation over months of I want this, but you can't do that. There's a lot of give and take. So why we created a WYSIWYG, ⁓ kind of what you see is what you get, is the designer now has ex the access to be able to drop in a ⁓ a th 3D file, be able to upload, you could use Rhino, create your curves, your design tools. You can then populate in our software all of your structures and designs, pull that into ⁓ you know, into the platform, including not so I'm talking about shoes now. You can you can also then pull in outsoles and actually have a full 3D design of your shoe. ⁓ Then export you can you can then export it out if you wanted to put it into VizCom or Nano Banana, ⁓ if you wanted to create a high fidelity 3D model or an animated model. Or you can effectively export out the file and plug it into a machine and knit it. I mean there's a lot of blockers. I mean there's so many blockers in the ⁓ design textile industry. So for those Those who are listening that are familiar with it, I mean, I could really quickly run you through how a product you go from i idea to finished product. I can talk about a shoe because we're doing quite a few of those right now. ⁓ you know, or even apparel, but you need a you effectively take a sketch or you take a three D model that you design, ⁓ like an a a visualization. Doesn't mean what you sketched or what you designed in Clow or some visualization three D model tool, you actually can make that. Right, because certain structures, which are different types of fabrics that you can make off the machine, don't play well with others. They don't actually go around the product the way that you may visualize or want. There may be pixelation. There may be, I mean, you may actually not be even able to make that product, right? But there's so there are there are limitations when you when you are, you know, 3D knitting a product or knitting something to shape. Then further down the line, you need a pattern maker, right? You effectively need someone to take your design, wrap ⁓ a last. So today you tape it up or you use other types of software to create the pattern. Then you give that to the engineer. The engineer then spends a week or two weeks trying to program that thing, setting up the machine, trying to get your ideas close to what you want, but it's never what you actually want. So when you get it, and you ask any designer, what you get is not what you gave them. And that's the worst, right? As a designer, I I was a designer, right? ⁓ when I handed off something to my pattern maker and I got my sample back and it didn't look anything like what I actually had like sketched, designed out in the tech pack, it was really disappointing. So now you start from, you're not starting from let's make it better. You're starting from how how can I get it closer to my original design? And that's what takes months. That's what all the waste comes in. Time, money. ⁓ that's why a lot of products end up dying, you know, during this development cycle. And that's why a shoe and all these big brands take You know, so anywhere from six to eighteen months, ⁓ to get, you know, to get through the whole process. A lot of that, the majority of that is in that in that beginning stage. So we say we're the Figma ⁓ for designing towards that as a kind of a North Star, is we're creating a single a single point of truth where all your like anyone can make ⁓ anyone can make design edits, changes, notations. in a single file format. And in that single file format, then you can export that out and you can create versioning history. So Ben, you may say, right, you're the merchandising, you may be in merchandising, I may be design. Someone may be in costing, right? You just are on the manufacturing side. They can make updates and changes and it updates and creates versions of that product. So you have V1, V2, V3. Right now there it sounds crazy, but you don't have that. Whenever I create a knit new knit program, I need to save it. I need to then create my own file naming system. in a ⁓ you know on my server. And then the only way for me to actually know what I I saved, if I can't remember, say it's a year or two later, right? ⁓ and I wanted to go back to a a shoe that I designed or a garment I designed, I actually have to spool up the machine. I have to knit it down. I gotta look at it. There's no visual at this it's it's very disjointed right now. So all we're doing is bringing all these things together into one place. So you can see it, make decisions, and and by the way You can actually see a high fidelity 3D model. And we're building our own visualization tool for knit and textiles right now. Cause just there's nothing out there that that exists. ⁓ so that you can actually see a thumbnail and then pull up the actual image of what that product looks like. And you can see who made changes when and what it looks like. ⁓ and we're building a lot of what we're building is coming from direct feedback from partners that we're working with. So we're not just building into a into a vacuum or a void. It's like, what do you need? I think I know what you need, you know. Let it and and then we're building and building these tools in collaboration. But ultimately is is creating the ultimate collabor collaborative tool because at the end of the day, that that creates speed and that creates ⁓ a product that everyone wants to bring to market faster. And at the end of the it's a designer ultimately guides a lot of that. They don't like it. You y you kill the you you kill the product. Or they may love it, it may cost too much to make. So therefore, then you kill the product too. speaker-0: I think I think people, even people who are in the industry, they they under acknowledge just how much of the typical product journey is made up of reconciliation. Right. We're made up of different different people pulling in different forces, pulling in different directions, getting those different work streams together in one room and figuring out how to make them make sense together. And it's it's too simplistic if you call it creative versus commercial tension or what have you. But like I've been surprised when I've worked with brands in the past. I've I I've sort of said that that so much is happening here that is not consistent with a single reference frame. I think that was also the promise of a lot of digital product creation in like Wovens and and and and other like apparel sort of areas was that what you need therefore is a digital representation of something that is or should be, or ideally should be, accurate to the thing that's actually being made. Now that's a different prospect in Wovens than it is in NIT because it's not driven the same way and and and so on and so forth. speaker-1: But y you know, Ben, and to that point, I I think I didn't realize this either until I went to go visit w probably one of the most advanced footwear factories in Asia. There it's still a very manual assembly process. There are there were thirty-five people on a line putting together a shoe. I couldn't believe it. You know, so it it took a bit of like you gotta rethink. I mean, if we want to onshore stuff or do things locally or be able to, you know, create new products that are more sustainable, you gotta rethink how you actually put put that that product together. And that's what we've done. is is it's it's forced us to think outside the box. You know, if I you look at like a software platform, I keep going back to Clow because it is a great, it's an amazing platform. But what they've done is they've they've created a a digital tool for traditional manufacturing. We s we step back and said, that's fine. We can be a piece of the manufacturing system, but what if we rethought the whole thing and and what are all where are all the blockers? Where are all the things that that are frustrating? that take time, that waste money, that kill an idea from the get, or, you know, which is even what which is worse, it's three quarters, when you're almost at the on the 11th hour about about to finish the product, the product dies because it's just too expensive. You can't make a revision. So we just thought, okay, let's let's rethink this thing. That's also the other thing is is if you make one, I mean this sounds crazy, right? This should be, it seems like it should be the norm. If you make a decision, it should cascade and waterfall all the way down. Instead of like today, you make one decision anywhere in the process, you gotta change it four or five times in your tech pack to your manufacturer. I mean, it's like it and and that's where that's where mistakes happen. That's where, you know, that's where time's wasted. That's where products get all goofed up. It's there's just no there's no cohesion between, you know, there's no cohesion or link between ⁓ a change in the process right now. speaker-0: I think what we've talked about a lot there is is like internal alignment and internal collaboration and internal communication. Let's let's engage with the external side of it, right? So the we talked about the Figma comparison aiming for ubiquity and democratization and accessibility, right? Let's let's pick up the bit from the quote before, which was the any nitty machine anywhere in the world side of things. Walk me through what you think the process of industrialization and drop in. swap out whatever you want to call it, ⁓ kind of integration into the supply chain, into the existing upstream infrastructure looks like. So when when we've we let's say we've worked in the browser, we've come up with a machine ready file that was born in ⁓ in in in CAD and a web browser, what needs to happen on the manufacturer's end to make that workable? And by extension, what would it take to take the setup you have at the ranch you talked about earlier? make that a viable thing to scale out further in California, or like you said in the Cotswolds, or anywhere else in in Europe, UK, Asia and so on. It so having anybody be able to design a net anything from their browser, that's a cool vision. And I think we've got a software part of it pretty well sewn up from the conversation we've just had. But there's more than software to stand up there. speaker-1: Absolutely. And and I I have to be we have to be I have to go in the size wide open, right? We there we have a lot of partners. There are people that we we what we're what we're proposing is the future, right? And what we'd like everyone to adopt. We really I have to understand there's still a lot of network engineers that that can help accelerate ⁓ what we're building as well. ⁓ so standing it up on the manufacturing side, what we can give what we're giving right now is what I say any any machine anywhere in the world, ⁓ Today we're focusing on, you know, with today we're focusing on Shima and Steiger and Shishane, right? Because that's the bulk of flatbed knitting right now. We paused development on stole because of what was happening with the company, right? It just didn't make any sense till everything shakes out and we see where it where it ends up. ⁓ but the majority of products will be made on those three types of machines. ⁓ it's a it's a we've created our own knit code called knit block that we can abstract into our our own ⁓ knit language. It you effectively can you can code it like ⁓ like JavaScript. ⁓ it's really interesting how you how we're how we're building this out again to make it more accessible. ⁓ what we're giving we're giving manufacturers now or effectively designers and brands is the expanded DAT files for ⁓ like a Shima. ⁓ an engineer can go in, make modifications if they want, ⁓ but effectively then take that and put it into production, right? Because there there is still a There is still a process that a product needs to go through. And we're we're not, you know I I know they're they're going to there's gonna it's gonna take time for everyone to adopt this new way of manufacturing. We need to prove it here ⁓ in Malibu. ⁓ and then we're talking with another we're talking with another group that's looking at standing up ⁓ responsive manufacturing here in LA, ⁓ or the US, LA and Portland, ⁓ that will be adopting the software. ⁓ and then as we show it and prove it, ⁓ how we can go direct to machine, do all the grading, everything, and how much more efficient it is. Then the and and and further, rethinking how you actually make products. So instead of garments coming in five or six pieces and then having to be assembled together, with our software, you actually can make it in one piece with two seam lines. You can make a shoe as a full 360 booty instead of having to sew on a, you know, create a horseshoe and then sew on the the strobe do a stroble and all those other components. There's other ways of that we're creating products, but that does change how you manufacture. And a lot of these Asian manufacturers who are, you know, we work very closely with have a setup where they've invested hundreds of millions of dollars into an assembly line and are there there is a little bit of pushback to want to change that overseas. There is an opportunity to change that here domestically because ⁓ we're we're effectively starting or building something from the ground. speaker-0: for me, with any anything that is sufficiently disruptive or is poised a you know, ⁓ the posture is sufficiently disruptive for something, I'm the p the part I always want to press on is what will it take to put it into the infrastructure that already exists. And I think you have a good answer to that, which is to say you need to prove it out by developing new infrastructure in country domestically, so that the benefits are more visible and then companies that are heavily entrenched and have a lot of inertia around traditional assembly line production, then start to see that it's not just an experiment. It's something that th they would need to reorient their businesses around. speaker-1: And and at the end of the day, it'll be faster, more efficient, and and you know, it'll be faster and more efficient, allow them to actually create diversification ⁓ with the machines that they have currently in stock. I mean, right now, which is kind of crazy, Ben, is you set up you go to a factory, it's set up only for garments. But that garment machine could actually make shoe uppers, could actually make, you know, furniture covers, could make automotive seat covers, could make medical products. It it is in in all if these machines are so adaptable. ⁓ w we should think outside the box instead of categorizing like a you know, just not to get super technical well, like a CMS five thirty ⁓ from Stoll, which is a your traditional flatbed knitting machine for shoe uppers. Well, now that now that we've reengineered how you can make a legging on it, we can take a very narrow needle bed, so like a very short width ⁓ machine, and now make a legging on it because we've redesigned re engineered how you make that legging and made it make it better. So now you now your machine can make more than just one product, which is a shoe upper. Now, as a manufacturer, you can now bring in, you know, you don't have those lulls in production. ⁓ but you know, a a big thing is I do have to be I I I was a little I was a little naive. ⁓ I mean, I made plenty of mistakes, ⁓ and I'm totally self aware on this. I was totally naive that everyone would be open arms and willing to adopt this. ⁓ I knew I would get some pushback from ⁓ knit engineers, right? Because there's a lot of skepticism. And then there's fear too that we're going to automate out their jobs, and that's just not the case. What we're doing is is we're taking I mean, what I thought was let like take take ninety percent of the lift of the the monotonous kind of making the base program. What if I flipped that or to the eighty twenty rule? Like, right? I took eighty percent of the the kind of the monotony of making the the base program and then only left you 10 to 20% of the time to innovate and engine actually engineer. Well, what if I flip that? I loop, did all the the heavy lift, got you 90% of the way there, and then allowed you as the engineer to then really push the boundaries of what you could do and and then reinvest 90% of your time into that. That's what we're talking about. And that's what we're doing. And that's where, that's where, you know, I had to change a bit of my I I had to be, you know. I had to ta I'd change a bit of my like my perspective on it after talking to a bunch of the engineers and realized no no, this is what we're doing. We're actually this is a it's it's like a it's a bit of a superpower, you know, it it's a bit of a superpower that allows you to do the fun stuff. speaker-0: Yeah. No, I think I think that's right. ⁓ now you you mentioned sustainability at the top as something that you're passionate about. Since then, we've said faster, better, more efficient a lot instead, ⁓ as the as the kind of wedge. Now, at last year's sourcing journal Sustainability Summit, I think you were interviewed by Kate Nishimura, who is a friend of mine. and as part of that summit, I I've I've watched the the interview back. You said you were finding it hard to sell the sustainability benefits benefits of everything we've just talked about, like, you know, reducing waste, doing things domestically and so on. And that you'd started purposefully treating sustainability as kind of the riders in the Trojan horse thing, and instead putting cost, time, logistics, and things first that have a secondary benefit in sustainability terms. Do you still feel that way? Do you still talk that way? ⁓ and do does that sentiment have anything to do with? tech and innovation budgets being reassigned away from sustainability initiatives? Or do you think it's just a case that brands respond far more readily to something that directly hits the bottom line? speaker-1: I think the latter. I think brands I just anybody, right? It's again being really passionate about sustainability. ⁓ I mean, we even hired a a PR company that was all I mean, really put us on kind of the sustainability s sustainability ⁓ like on on that path of ⁓ being kind of a ⁓ a thought leader in that space. But I realize it didn't really move the needle. ⁓ what would move the needle was and what brands do care about because it's a business at the end of the day. Is how can I bring a product faster, cheaper, and more efficiently to market? How do I I mean on all these things are huge ⁓ in environmental wins, right? Can we can I quickly change a program? I mean, if I tell Jodie right now, if I change if I took a program, ⁓ if I took a if I if I design a program out of a polyester fiber and I want to change it to a wool, I have to start over right now. If I use traditional ⁓ traditional machine manufacturers CAD software. It's a redo. So most people don't do it, right? You just stick with it and I'm just gonna I'm gonna run this thing because I s I've invested a year's worth of time and a few hundred thousand dollars to develop this thing. I'm gonna run it all the way to the end. Well we give you that flexibility now. Now you can swap it out because we do something called stitch density, which tells you how, you know, how dense the the the knots are, effectively knots between each fabric. And then we take that calculation and instantly create a new program for you. So we're removing a lot of those barriers. But No, ⁓ it was it's it was very hard to to walk in. Every I love it because I l everyone's heart is in the right place. But at the end of the day, you have a everyone has a business and you have either shareholders or you've got, you know, you've got you just have to you have to pay your bills. I I'm I'm not naive that. And you know, the if I if so why why near shore? Well, because then you can be maybe it's a little more expensive or maybe not, right? Because everything that's going on with shipping lines and you know what happened with COVID, there are shipping disruptions and it is and when you factor in kind of all the costs of ⁓ lead time, ⁓ shipping costs, right? And and it's a bit volatile right now, ⁓ fuel costs, ⁓ you know, warehousing, distribution, you know, it almost it does create cost parity when we when we lay the whole thing out and you talk to the right people ⁓ within the company, they understand that it actually the time savings and the cost. are about roughly the same as if you're to do it overseas. We don't have the infrastructure currently here domestically, but there are ways to do this with, you know, 3D knitting that allows you to, you know, effectively have this the the same cost parity if you're to make it overseas in Asia. You have to rethink how you make the product. That's all. You can't have all the pieces. You can't have 35 people mainly putting together a shoe. It has to be a two step process. So it has to be a new type of shoe. It has to be a new type of product, right? Same thing with a legging. It can't be like a Lululemon or an Aloe legging, right? It can't be thirty five different pieces with seam lines everywhere. No. You should actually engineer all of that into your legging. Couldn't do that before because it was too expensive to pay an engineer to do that. ⁓ but now we're we're we're removing that barrier to entry. speaker-0: Fine. And I I I I buy that. I think I think I agree with the fact that and I've heard I've heard brands talk about it, to be fair. I've behind closed doors, you will hear companies say sustainability is a not a marketing term, that's being a bit uncharitable. ⁓ sustainability doesn't directly drive behavior and things the way that that you would think, ⁓ and instead what they lean into is. How do they become faster? How do they become more profitable? And the sustainability benefit is an upshot of that. ⁓ so yeah, I think I think I agree with that. speaker-1: Just just to like kind of follow on that is at the end of the day, the consumer the consumer is the driver of what a what a business is going to provide. I mean, if if sustainability if if sustainability was I mean, I really like kind of ⁓ for a majority of people, the the forefront of what everyone wanted, then we probably wouldn't have the HMs, Zara's, the, you know, everyone wants you can't afford it, right? It's quite quite expensive. ⁓ you know, so there there there's businesses. supplying to consumers that want something. What I'd like to do, and what we're truly I mean, call me a like, I guess I I I'm a bit of an idealist, but what I'd like to do is I think everyone's got their heart in the right place. All brands, every brand. I don't think any brand really does it, right? They want to do the best and they want to serve their customer the best. That's that's how you have a you build a successful business. You know, maybe that's where I get from the hosp my hospitality side of my that's where I, you know, from the the hospitality side of the ⁓ my upbringing is is, you know, you listen to your customer and you know, deliver deliver more more value more v you know value than what you're charging for is is if we could, you know, if if we can create accessibility in this space and then other spaces across the line to every brand and and give them the option where onshoring or localizate localized shoring ⁓ local ⁓ local manufacturing or sourcing more sustainable fibers is ⁓ is has some cost parity, then I think everyone would do that. I don't I don't think that people would intentionally say, no, I want to use you know, really, you know, resource intensive or pollutive, you know, ⁓ you know, ⁓ you know, just ⁓ fibers that don't that aren't great for the environment. I think people would, including brands, designers, they would all choose to kind of go in that direction. It's just it's very it's very tough right now. speaker-0: Yeah. No, I I I would a hundred percent agree with that. And I think I've I've I've come under some fire for before. But you know, people, particularly in Europe, are very up in arms about Sheehan and the ultra fast fashion side of things. And and I get it. I understand it's a it's a extractive business model and and what have you. They're making what people want to some extent. They they they are making what people are willing to buy. ⁓ and I think it requires you to to do something different and demonstrate that thing. in order to showcase that there is an alternative. So I think sorry, I think that's right. Now, we're a fashion technology publication, but power and footwear are just two of the few verticals that it seems like you want to go after with loop. and you've you've you've I think you've already hinted at this. Like you know, if a knitting machine can do one thing, it can be pressed into service to do multiple things. Everything from Interior design textiles to medical compression wear, and maybe even the knitted skins that you see over humanoid robots, ⁓ which sounds far-fetched, but clearly somebody is making that ⁓ that material. fashion's got a pretty long history of asking/slash demanding industry agnostic platforms to adjust what they do to fit fashion's idiosyncrasies and priorities. So I'm curious first, how big of a fixture you think. Fashion is going to be on your roadmap going forward. And fashion I'm using is a catch-all term for apparel footwear, ⁓ accessories and so on. and second, whether you see loop as a platform that just works for every use case without adjustment and that the principles of of 3D Knit are the same for everything, or if you think it's something that needs to be progressively verticalized in order to realize its full value. speaker-1: So ⁓ you know, fashion is our focus and and that's what we started off, ⁓ really focusing in on. Footwear, we've we've kind of ⁓ we've kind of shifted our focus into footwear because the complexity of footwear we realized the the tolerance, the allowable tolerance to that y and the accuracy has to be near perfect. And the the complexity of footwear actually covers a multitude of different industries. And like we talked about earlier, these knitting machines, it's not a an apparel knitting machine, like a sweater machine. It's not a footwear machine. It's not a ⁓ you know, it's not a ⁓ furniture machine, right? And that's what everyone loves to call it, right? It's like this is what it does, this is all it does. Well, no, actually, if you if you step back, look at it a little bit differently, that one machine can actually, you know, go across ⁓ produce products across a multitude of different ⁓ industries. And we're doing that today. We're working in humanoid robotics. we're working in the medical space, we're working in footwear, ⁓ automotive, ⁓ apparel. ⁓ I mean, it's like ⁓ we did some stuff in aerospace. ⁓ I mean there's all kinds I mean it's the list goes on and on. And Ben, we're using the same machine. And and in some cases, we're using the same fibers, right? ⁓ to to be able to create these products. So it's all about, you know. how do you set the machine up is how you can actually leverage the machine. A lot of it the the the kind of the the the linchpin of how we do this is using our varying three D loop platform to be able to do this. But it's ⁓ it's kind of just rethinking rethinking the ⁓ the the business as overall. I mean that's why again, I think domestic manufacturing is going to possible because now one machine, you're not you're not stuck waiting for an order, right? ⁓ I'm, you know, do you have seasonality to orders or lead times. Now this machine could really be running 24 hours a day, seven days a week. Well, maybe not 24 hours. You got there's always a couple hours of downtime. ⁓ but you know you could you could maximize the efficiency of these machines. Therefore, you could amortize that cost, making it more affordable for manufacturers, making products across the board cheaper. And then further, we're taking so does apparel does apparel lead design in in automotive or does automotive now inspire apparel and footwear? Now we're taking all these inner like cross cross disciplinaries and we're we're allowing them to kind of ⁓ feed off each other. speaker-0: ⁓ okay. So we talked about democratization. Who do you see as your long-term customer though? Because I think that the there's the idealism point that you talked about, which is making it available to everybody and creating the Figma for textiles, lowering barriers in software and so on. it's easier to do that, I think to put together a potentially massive addressable market through a browser than it is to make to do the kind of distributed domestic production side of things and make it more viable on demand. Who in the long term, who are you selling to? Do you think the brands are the lever for making this kind of change? Do you think it's their factory partners? Do you think it's d whoever's setting up this new infrastructure for domestic production? speaker-1: So it's ⁓ it's brands are gonna lead this. ⁓ and we're working with kind of all the major brands right now. ⁓ brands are gonna lead this. ⁓ designers are are actually who we're focusing in on. And that's ⁓ that's ⁓ that's kind of our that's who we've been building the the software for, is giving designers the accessibility. ⁓ and at at you know, with a click of, you know, either it's computational design or just traditional designers. We do need to teach traditional designers how to use some of these, you know, computational design tools. But I ⁓ but everyone but ⁓ not everyone, but I would say the majority of new designers coming in the industry are are really very, fairly familiar with kind of these new computational design tools or CAD tools that are coming out. And you can now we're we're not trying to be everything for everybody, right? Like ⁓ we're focusing on designers, we're allowing them to use third party CAD tools, pull those those three D designs into our software, then you work within our software, then you can export it out. into all these different into these different types of machines to make samples immediately. You know, our kind of the as we as we go forward, it I would like to it the goal is have it trickle down. Right. So brands will lead this. They'll may they'll we're pushing this into manufacturers like large global manufacturers right now. That allows us to open up or rethink how we manufacture products, even onshore products. Then we'd like to I'd like to tear it down to smaller brands and then ultimately allow anybody to be able to come into the space and create design and make products ⁓ and launch their own brands. This is to me, this is like the you know, it's like the nexus of ⁓ what ⁓ DCG printing or embroidery. Like this is the this is the ultimate abstraction of that, right? Whereas instead of taking a blank that someone made and then creating putting a graphic on it, what if you can make the whole garment with your graphic, with your design, with all the with all the attributes of, you know, ⁓ What kind of the physical attributes ⁓ of that textile and be able to and then effectively be able to stick it on your own knitting machine and then knit it out. speaker-0: Okay, so when we think about the letting anybody design anything democratization point of view, ⁓ you talked about computational design a few times, which I think is is probably right to say those tools are still primarily aimed at people with formal design education. Then when you think about actual kind of unit one director garment, digital print sort of workflows where literally anybody can design anything and and get it on a t-shirt, get it on a hoodie or so on. w it's a very interesting time to be talking about letting anybody make anything they can imagine because there are an awful lot of tools in people's hands from a generative point of view that allow people to imagine an awful lot. What do you see happening there? Do you see generative kind of algorithmic do you see that being more of a force in that kind of future design that you're talking about? speaker-1: I mean, that currently that's what well, that's what we built is we have our own algorithms that are able to abstract these these designs or abstractions of what people are creating. ⁓ we're already we're in a lot of ways proving that ⁓ we're lowering or we're lowering the barrier to entry into the space. So traditional ⁓ network engineers don't actually need to come in to create a product right now. We're working with just designers who don't have a have any network ⁓ background or engineering background. And they're creating designing, creating, knitting off products in real time. So it's ⁓ you know, there there are there are other things that we're looking at. We're using, you know, we're using Claude to to create knit structures. ⁓ for engineers internally are doing that. It's really it's really interesting in populating ⁓ our back end internal system. ⁓ we're leveraging all kinds of different tools to to do this. I mean ultimately, you know, I think so why why do we have like tools like Viscom or Clow or you know our people are creating these amazing abstractions or 3D abstractions of designs is because it's a to me it's a substitute for the physical product, right? You want to show off to, you know, leadership. Hey, this is the idea, this is what it's gonna look like, how amazing. And then they say, okay, yeah, now go make it. Well, what if I could tell you the same amount of time it takes you to make that, you know, VizCom high high fidelity 3D model that I could actually export out and give you the the real thing in your hands. at the same time. And that's what we're doing. speaker-0: Fine. And finally, what do you see as the timeline for this? Like realistically, what do you expect apparel slash footwear production to look like in twenty thirty? And so that that's one of those dates that feels really abstract and felt like a long way off for a long time, but we're halfway through twenty twenty six. So it's coming coming down the tracks pretty quickly. Yeah. Can do you think do you think the industry setup that you're talking about can meaningfully change in the next few years, or are we looking at a longer timeline for turning things around here? speaker-1: I do. ⁓ I I do and we're already kind of proving that out right now. We're doing it with a few partners on a ⁓ kind of a beta test model. That's why we haven't ⁓ over overly opened up ⁓ Luke ⁓ to a broader ⁓ kind of ⁓ user base. ⁓ there there's some conversations of you know being able to prove this out in a kind of high velocity, like high velocity way with a couple of big brands that really understand the space so we can build out the infrastructure correctly. But I do think in three and a half years, the you know, the the landscape of how you make products, it won't be ⁓ it's not gonna be what we like the dream of anyone could just log on with their iPad, you know, log in a loop, design something, you know, print it out onto to their to their own like on their own knitting machine. I don't think it'll be there yet, right? But ⁓ what it'll where where it'll be is I think it'll take we'll be able to take ⁓ ideas. What we're doing right now, I mean today. One designer with no knit engineering background. ⁓ we we did this in ⁓ in a work session internally with ⁓ one of our computational designers that has no idea about knitting. ⁓ she created ⁓ 12 different variations of a shoe upper. This would take years, years, hundreds of thousands of dollars right now in the current in the current system to be able to do this. I mean, we're we're currently proving this out internally. We're doing this with partners, ⁓ you know, you know, partner brands. ⁓ and we're now deploying this into their ecosystem. So ⁓ you know, I think in twenty thirty you'll be able to you'll see generative design products ⁓ coming out much faster. And I think you I think you'll see a lot more ⁓ co-collaboration or ⁓ inspiration coming from ⁓ the anywhere from architecture, ⁓ automotive, footwear, apparel. You see a lot of this crossover ⁓ you know, in design, ⁓ kind of effectively feeding on each other. and quite frankly, I I think it'll come all from ⁓ the loop platform that we're building. speaker-0: Well, let's aim to sit down again in twenty thirty and ⁓ see how this way. Maybe not maybe not that long, but at the very least we can we can pencil it in for that. For now though, ⁓ Garrett, I appreciate this conversation. That's ⁓ insightful from my side and ⁓ network not something I spend a lot of time immersed in, but would like to. So this has been a good ⁓ crash course for me as well. And I'm sure the listeners have got a lot out of it. Thanks very much for joining. speaker-1: Thank you. Thank you, Ben. And ⁓ pleasure. Love being on your show. speaker-0: And that's the end of my conversation with Garrett. Hopefully it's given you some insight into knitting, if you don't already spend your days immersed in it. And if you do, then I'd be keen to hear what you think about the viability of what Varian3D is trying to do. I like doing these weekly prestige interviews with CEOs because for me, they're an opportunity to kind of press people on what the practical implications of their visions are, irrespective of whether those visions are confined to Selling software that's generally available and that caters to a narrow category, or whether they've got like whole industry transformation scope and ambition in their sites. I'll have another one of these for you next week, the week after that, and so on. So while we're still sounding out and feeling out new ways to talk about industry news, which you might have already listened to by the time we hear this, my promise to you is that you'll be able to come back here every week and listen to me quiz someone senior in the world of fashion technology for the foreseeable future. So until next week. Thanks for listening and I'll talk to you again really soon.