The Lattice (Official 3DHEALS Podcast)
Welcome to the Lattice podcast, the official podcast for 3DHEALS. This is where you will find fun but in-depth conversations (by founder Jenny Chen) with technological game-changers, creative minds, entrepreneurs, rule-breakers, and more. The conversations focus on using 3D technologies, like 3D printing and bioprinting, AR/VR, and in silico simulation, to reinvent healthcare and life sciences. This podcast will include AMA (Ask Me Anything) sessions, interviews, select past virtual event recordings, and other direct engagements with our Tribe.
While there is no rule for our podcast content, the only rule we follow is to provide our listeners with a maximized return on their attention and time investment.
Follow us on Facebook, Twitter, and Instagram @3dheals, and check out the links in the show notes.
3DHEALS Links: https://linktr.ee/3dheals
🛑 Disclaimer
The content of this podcast is for informational and educational purposes only and does not constitute medical, legal, or financial advice. The views and opinions expressed by the host and guests are their own and do not necessarily reflect those of their employers, affiliates, or any associated organizations.
While we discuss emerging technologies in healthcare and 3D printing, listeners should consult qualified professionals before making decisions based on the information shared. The mention of specific companies, products, or technologies does not imply endorsement.
This podcast may reference early-stage innovations and concepts that are not yet FDA-approved or commercially available. Always follow regulatory guidelines and ethical standards when applying new technologies in clinical or professional settings.
Would you like this disclaimer read aloud at the beginning of each episode, or placed in the show notes?
The Lattice (Official 3DHEALS Podcast)
Episode #93 | 3D Design & Geometric Intelligence with Elissa Ross
What happens when advanced mathematics meets manufacturing? The result is a new way of creating products that range from record-breaking running shoes to life-changing medical devices.
In this episode, we sit down with Elissa Ross, mathematician and CEO of Metafold 3D, to explore how her company is using mathematics to reshape design and manufacturing. Metafold’s platform is built on geometric intelligence which is her company's novel approach that transforms shapes into data that can be analyzed and improved. Instead of simply representing objects the way traditional CAD software does this method allows engineers to simulate and optimize designs with greater speed and accuracy.
At the core of this approach is implicit modeling with signed distance functions. While it may sound highly abstract, it has very practical applications. This technique allows manufacturers can run thousands of simulations in the time it would normally take to analyze a single design. The result is shorter development cycles and products that perform better in the real world.
Ross also reflects on her journey building Metafold, from its early focus on 3D printing to its current role serving major industries. She explains how their API-first platform gives customers the flexibility to solve specific challenges, such as analyzing tolerances, reusing similar parts, or predicting how designs will hold up under different conditions.
Whether you are a designer, an engineer, or simply curious about how mathematics is shaping the technologies around us, this episode offers a thoughtful look at the future of manufacturing intelligence.
Shownotes: Check back soon
Podcast Engineer: Faith Fernandes
Subscribe to our premium version and support the show.
Follow us:
Twitter
Instagram
Linkedin
3DHEALS Website
Facebook
Facebook Group
Youtube channel
About Pitch3D
Hello there, welcome to the Lattice podcast, episode number 93. Today, we explore the world where mathematics meets manufacturing innovation with Elisa Ross, co-founder and CEO of Toronto-based startup Metafo 3D. Elisa and her team are transforming the way industries analyze and understand 3D geometry, using advanced geometric intelligence and implicit modeling. Whether you are a designer, engineer or anyone fascinated by the creative side of math, get ready for an eye-opening conversation about shape and the future of making things. Please listen to the disclaimer at the end of this podcast. Please listen to the disclaimer at the end of this podcast. Hello, hello. Thanks for joining the pod.
Speaker 1:Today we have the CEO and co-founder of Metaphone 3D, elisa Ross. Hello Well, elisa. I feel like today's interview is a bit like a challenge for me because, even though I love math, this is like one of those mathematical challenge I encountered when I was in school. It's both exciting and kind of fascinating, but also I understand the limitation of my knowledge. So first of all, I'd love to just hear a little bit about what Metafold 3D is and a little bit about yourself what Metafold 3D is and a little bit about yourself.
Speaker 2:Yeah, absolutely Happy to Well. First of all, thank you, jenny, for the invitation. I'm super excited to be here and I promise we won't go too deep on any mathematical topics. Metafold 3D is a startup. We're about five years old now, we're based in Toronto and we're all about math. As you've already introduced, I'm a mathematician by training, specifically geometry, and we can get into that if you like. But Metafold is all about bringing geometric insight to manufacturing, and our product is called Geometric Intelligence, and it's really all about turning geometry into information that can be used to make decisions. We can unpack all of that, but that's the highest level.
Speaker 1:Awesome, I love it. Yeah, I think we want to unpack some of the maybe the most important concepts of your software, because I am also learning as we are doing this interview, and why don't you unpack some of the most important concepts at the core of your software?
Speaker 2:Sure. So geometric intelligence is our technology and it's really a platform for shape intelligence, to shape information and analysis. So what I mean by that is that we take geometry as input and we give people using our software information about that geometry. So why would you want this? The idea is really that you know, shapes are not that meaningful on their own when we manufacture things. These shapes exist in the world to perform a purpose, to serve a purpose, and so our approach is all about helping people understand how their shapes would perform or how they will be manufactured in the world before they go ahead and make those things.
Speaker 2:So, very specifically, we convert geometry into information about performance. Like you know, how well will it perform under certain loading conditions? Let's say, we convert it into manufacturing information. We also do a similarity. We can look at similarities, so answer questions like have I made something like this in the past? And we can also address things like tolerances, which have a huge impact on manufacturing at scale. So, again, we can get into all of those topics, but the thing I really want to focus on or emphasize is that we convert geometry into information that can be used to make decisions Under the hood, of course, and we can get into the mathematics of it under the hood. We take a fairly different approach from traditional CAD software.
Speaker 1:Yeah, I want to unpack the concept of CAD before because I had the same misconception about it, since I am not a designer and I don't work with it day in, day out. But CAD software itself does include analytics and also provides some kind of analysis for manufacturing and outcome. I mean, some of the bigger companies that I have in mind are Autodesk or Synopsys Synopsys sorry they are. They also have design capability and analytics capability. Where do you guys are you? Are you guys part of this chain of information or you are completely different from that?
Speaker 2:Yeah, you're right, all of these platforms do provide some kind of analysis capability, and the reason is what I stated before, that the shape is kind of useless on its own.
Speaker 2:You need to know something about the shape.
Speaker 2:Where we really differ from those platforms is in the way we represent geometry, and that goes into the details of our technology. But, in short, we represent geometry using an implicit approach which is different from the way traditional CAD is represented, and this allows us to extract different kinds of information and do things differently. So it means that we have, for instance, a faster path to simulation, and so if you're doing simulations one at a time, then conventional CAD is probably a great choice. If you're trying to do a thousand simulations at once, then maybe we should consider having a conversation. This is the kind of thing we can help with. So that's actually. Another component of our technology is that we are really built around working with large volumes of 3D data. So if you have a lot of geometry, you're trying to do a lot of experimentation or you're going to produce a lot of parts, then this is a very good spot for our technology to slot in, because if you use conventional CAD, you're probably looking at those things one at a time, which is very time-consuming.
Speaker 1:And yeah, I have to say this is a very enlightening conversation for me personally, because decades ago, believe it or not, I actually did research in simulation of a 3D shape which is aneurysm, because we want to use a mesh to predict what kind of aneurysm shape would rupture.
Speaker 1:And actually that kind of information now translated to a company called the HeartFlow and a bunch of other medical-related simulation tools to predict heart attacks and aneurysm, ru related simulation tools to predict, you know, heart attacks and aneurysm rupturing medical device, but they're all based on the traditional CAD. Now I thinking, thinking back, and I remember I have to wait for a whole day for the analysis to come in, for the stress sheer stress calculation and, yes, just to create the mesh itself takes a couple hours and we're talking about circa 2010. And I remember that I work with PhD, so like I will like figure out the shape of the aneurysm to be simulated, and then he would keep the computer on overnight to generate something the next day. It's either a fail or success. It's tremendous amount of work, but the results you know. It just takes so long. I remember that had we had implicit modeling capability, things would be different and maybe we can test 10,000 different aneurysm models.
Speaker 2:This is possible. Yeah, I mean it's also possible that that approach was actually great and that it was just fine In some cases, depending on the type of simulation you're running, and then they need to change the format of the data into a different format for simulation. Maybe it's a different person, that's what you described. There's someone else who actually runs the simulation and so, you know, in large companies this might be a fully separate team. So if you're trying to iterate on a design and you're doing that super slow process, this just takes too much time. What we can offer is helping people bring a lot more of that information to the designer really actually in some cases so that they can run their simulations and get information that can inform their design process much more immediately.
Speaker 1:So are there certain type of shapes in particularly would benefit from Metafo's software.
Speaker 2:I wouldn't say there are certain types of shapes. No, I think what's kind of interesting to understand is that so much of the way we interact with geometry on a computer will come down to how you represent geometry. And so for us it's not so much that there's certain shapes that are more amenable to approach than others, it's more kind of levels of accuracy, let's put it like this, and depending on what you're trying to achieve, you might choose different representations. So for manufacturing, you probably need a traditional representation like a B-rep or even like a test-related format, like the STL probably familiar to you in the 3D printing world. But for what?
Speaker 2:Our perspective at Modifold is that if you're trying to get information about the shape, it can be very helpful to move to an implicit representation because you have all this information about the shape. It can be very helpful to move to an implicit representation because you have all this information about the volume of the structure and you can do operations on it that give you information that you can't get with a traditional format. So it's not so much about the type of shape as it is maybe in the kind of information you're trying to extract from your geometry.
Speaker 1:Alyssa, I view you as my mathematic teacher, so there are two concepts I'm hoping that we can unpack for average people who listen to our podcast. One is implicit modeling, which is a new concept to me, and the other one is sign distance function. I know they sound like big words, but actually in reality, after a while, I feel like they're actually not that complicated to understand.
Speaker 2:Yes, I agree, and actually they're they're really one in the same in a lot of ways. So let's start with the signed distance function. This is a very um, simple mathematical function and it it operates on three-dimensional space. So you begin with a shape, a surface in three-dimensional shape space rather, and the sine distance function is defined by that shape and for every point in three-dimensional space it returns the distance from that point to the surface of the shape. So this is a function that is zero, I would say. In mathematical speaking. We would say identically zero at the surface of the shape. Everywhere else it is non-zero and is actually equal to the distance from the shape. If you're outside the shape, it's positive, if you're inside, the convention is that it is negative.
Speaker 2:So already this gives us such a rich space of information that's not present in, let's say, a triangle mesh. So for comparison, a triangle mesh is a set of triangles that you stitch together to approximate the surface of a geometry. Important to understand in both cases. This is not geometry, it's a representation of geometry, and I like to make that point because people often have kind of a mental idea of what geometry is, what it looks like, and that is always approximated on a computer. So the sine distance function is a precise and complete definition of the shape. But when we represent it on a computer we probably need to sample a grid to, like you know, sample the distance across the grid of values to really understand where that shape is. So that's the concept of sign distance function, Clear so far.
Speaker 1:Yes, so the mesh and the sign distance function. Would it be accurate to say they're both kind of ways, different ways to represent a 3D geometry?
Speaker 2:Absolutely, absolutely. And there's a third way. There's multiple ways, but a third common way is the boundary representation format, the BREP, and this builds up surfaces through points, lines and subsurfaces basically too listeners.
Speaker 1:Well, so I want to give us some real-life examples, if you're allowed to, or maybe some hypothetical examples, of how Metafold works, as opposed to, let's say, autocad and all the other CAD software that has some kind of intrinsic analytic capability, and also for people who are already working with those kind of softwares can they use your program seamlessly, kind of integrated?
Speaker 2:into it. Yeah, these are. These are good questions, um, maybe just because I want to close the loop on this. You did ask about implicit modeling. Yeah, I want to say that implicit modeling is just refers to using signed distance fields, usually to build up more complex geometric shapes. So, like conventional modeling, you're usually, you know, in a boundary representation or creating a triangle mesh, and when you're doing implicit modeling, you're using some kind of volumetric representation, probably a signed distance field, and this is actually kind of a beautiful thing. If you think about signed distance fields, you can represent them as a grid of values. So if you, if you want to say, add two together, it's as simple as arithmetic on that grid. So certain operations become much simpler in implicit modeling world, which is why people tend to like it. So, from uh, from that to your question about real world applications. Yeah, so I'd love to give some applications and I, given your audiences is in the health space, I'll try and keep them kind of health care adjacent.
Speaker 2:Um, and actually, on that, one of the the big application areas for us is in performance sportswear, specifically footwear, so running shoes, and here the goal is to create the most performant running shoes possible. So we work with teams who are working on that problem, that problem. There's been many material developments over the past couple of years and also just developments in the way the materials and geometry are put together that have created pretty remarkable outcomes in terms of athletes, like we've seen the marathon records being broken and all sorts of other similar records being broken, largely by footwear technology. So there's a real interest in pushing forward this technology. The standard process for producing footwear is to come up with an idea, physically fabricate it and test it.
Speaker 2:Okay, so this is certainly a good way of getting information about geometry and material.
Speaker 2:However, it's very slow, so teams will spend months in these prototyping cycles trying to iterate and find the best performing shoe, but in fact, it's not an exhaustive search because that is just such a time-consuming process. So what we help with here is taking a broad sweep of geometry, taking a material palette that's potentially specific to that manufacturer, and then helping them scale up their experiments. So really simply, all we do is replicate their physical tests digitally and then radically scale them up so that they can simply get so much more information than they would through a physical or manual process. That makes total sense. Yeah, and that is. You could imagine this same process working in a more healthcare related field, where you're trying to really, you know, provide some level of cushioning and you want to make sure that you're achieving exactly the right response to force, depending on the application. But often we see the same combination of geometry and material. You want to vary both of them systematically so that you can understand how your part performs.
Speaker 1:Yeah, I mean, I can see this can definitely translate, not just in the shoe industry, but all kinds of other wearables or pathetics. What about implants? Would that be? If people want to iterate different designs for hip implants, for example, can they use your software to somehow scale up their design process?
Speaker 2:Yeah, absolutely. I mean, it's the same kind of thing. What is sort of interesting about the implant world, though, as I understand it, being an outsider to the space is that it is patient-specific, and so that element might indicate that you know you're probably for a given patient. You might not want to spend too much time, you know, physically testing, and that might actually make it a very good use case for a small digital exploration, like given the patient geometry. We're trying to find the best performing implant. Here's a systematic way of studying this. It takes, you know, an hour or whatever. It takes digitally, and then at the end of it you find the candidate that works the best and will yield the best outcomes.
Speaker 1:Well, as much as 3D printing industry would like to have personalized implants, the reality is the majority of the implants are not personalized, but more and more implants are now 3D printed. I think even just like one level up improvement in design not personalized, but let's say model certain X or YZ can be a better fitting for osteoporotic patient and stuff like that Like just incremental improvement. They need design iterations as well.
Speaker 2:Yeah, exactly, and that's a great point, that a lot of these are based on kind of best practices as opposed to lengthy iteration cycles to find the optimal solution, or not necessarily the optimal, but an optimal solution. So yeah, absolutely, that would be a great fit.
Speaker 1:And if I'm not?
Speaker 2:this speaks to kind of like a bigger theme for us as a startup, which is that we have really moved away from focusing on 3D printing. So we, you know, when we began, we away from focusing on 3D printing. So when we began, we were very focused on 3D printing because there are all kinds of interesting geometric things.
Speaker 2:Of course that's how we connected, but as a market it's just not quite growing the way that we believe it should and hope it will. But more than that, in conventional manufacturing, whether you're talking about medical or footwear or aerospace, there are so many interesting geometry problems and interesting places to optimize. So, yeah, but understanding performance is one of our key pillars of turning geometry into information. It's not the only one, so we also work in instance or on-demand manufacturing. We work in kind of similarity analysis for part reuse and also tolerancing. I don't know if any of those resonate for you, Denny, or want to pursue further.
Speaker 1:I want to go back to the fact that you pivoted away not away, but like expanded your horizon significantly because you discovered market and we frequently interview people who did that, because I think that's a natural progression of knowledge and insights as an entrepreneur. And it's actually good for you, because 3D printing industry is a small industry and it still is growing very fast. But to survive as a company, you need to discover new markets, you need to create the market, and I'm just surprised in a good way that you have found a much bigger market that is not well served by the current software providers.
Speaker 2:Yeah, this is. It's exciting for sure, and you highlight a bit of a point here, which is that the current software providers and CAD in general is a very old and sticky kind of set of technologies of geometry we were talking about earlier. They exist both because they are pretty good, but also just because they exist and they've got inertia and they've got people who've got decades of files in those formats. So it is interesting to be approaching this large market of manufacturing with a new approach to shape analysis and geometry understanding, and we do certainly believe that there's a lot to be gained from this approach that's distinct from what the kind of conventional players are doing.
Speaker 1:Yeah, and also I feel like the market probably is also slowly responding to you and your company's product because, like you said, they weren't aware of new ways of doing things because of the inertia. And how do you tell people you know this way is a better way or more productive for you?
Speaker 2:I mean, I think it's. I think that's been kind of a key understanding for us too, is that it's not really the message is not hey guys, this is the better way, it's more. What more can we do with geometry? Ultimately, everyone wants the same thing. They are using geometry as a tool to produce something, and the last thing they want to be doing is thinking about their geometry representation. They just want to design a thing, they want to understand how it's going to work and produce it, and then obviously, they want to make that very profitable. And so to make it profitable, you ultimately need more information, and so this understanding from us. It's not about changing the way people design things. It's about taking the things they're already doing and giving them more information about them so that they can make better things, make faster things, increase their yield. You know, stop redesigning the same part over and over all these kinds of things.
Speaker 1:I think that you're one of the companies that is cloud native, the next generation of software provider, and now you're also going API first strategy in terms of business development. And the fact is, it's not just geometry analysis. You also provide a suite of benefits like instant quoting, tolerance analysis, like you said, and smart part reuse. I'd love to hear you just unpack some of those features, because I don't see this very often.
Speaker 2:Yeah, for sure, and I think like a couple of things here. So, cloud native yes, we have a cloud offering. However, we are not exclusively cloud and this has been like another sort of semi-painful lesson to learn right that if there's certain industries, for example, aerospace and even, to some extent, sportswear, where they just like they don't want to be on the cloud, and I get it. So we have solutions for people who want to achieve dramatic scale using the cloud and we have solutions for people who want to run it on the computer under their desk. This is all possible. But yeah, so in terms of the other offerings we have, it is all rooted in the same technology. So using signed distance functions to gain understanding about shapes For the InstantCAD quotation super interesting problem.
Speaker 2:Quotation Super interesting problem. And I think increasingly we're seeing more and more manufacturing that people are trying to manufacture more domestically for all the geopolitical reasons. But really interesting problems come up in the on-demand manufacturing space. So the idea in that business is that those on-demand manufacturers they receive a part and they need to provide a quote as fast as possible. So they need to provide a quote to the customer on how much it'll cost the customer to buy, but they also need to have a supplier cost. They need to provide, you know, like put it out for their suppliers to produce for a certain price. So how do they get that information?
Speaker 2:Certainly, it's a function of things like material and lead time and quantity and a whole bunch of non-geometric factors, but at some point there's a lot of shape analysis that needs to come in there. So, understanding the part, what kind of machine do you need to make it on and what are the factors that would make it more or less complex to produce? How many hours would it take to produce this? So we don't do those things. What we do is we convert the geometry into a set of features, geometric features and other forms of data that those companies can then ingest and build their costing model. So I hope that makes sense. We're not providing the price, we're not providing the process. We're converting the geometry into other forms of information, other measurements, different kinds of analysis results that allow them to create models that are accurate for pricing.
Speaker 1:Does it include manufacturing process materials and that kind of information in your geometry outputs?
Speaker 2:So we are really strictly on the geometry side, so we provide the geometric features. They bring all the non-geometric features and that's why they are building the pricing model themselves. But we can, you know, via our API, just give them all of this data about the geometry, and what this means is that fewer parts where an expert needs to open up the file, just one at a time, spin it around, estimate the time it would take to produce it, which is kind of the state of things at this point. There's some automation and there's some expert quoting that has to go on. So how do we increase the amount of automation for that process?
Speaker 1:I definitely understand that you guys are purely in the geometry business. However, doesn't it sound lucrative if you are able to create an end-to-end pricing quote instantaneously?
Speaker 2:Let me tell you this is quite a hard problem. Yes, it is definitely lucrative. And there are businesses that are operating on this basis, for sure.
Speaker 1:I think I can imagine why some of the manufacturers would not want to share the data that they use because it's proprietary for their own process. Definitely, yeah. And also, rome is not building one day. Metaphode is not coming out of thin air, so I want to just go back to your original founder story, if that's possible, to tell us a little bit about your early days before Metaphone and how that translates into today.
Speaker 2:Sure, I don't know how far back you want to go.
Speaker 1:By the way, just curious, I was reading the history of cat, which I will share a link in a podcast. Are you related in any way with Doug Ross?
Speaker 2:I am not. I'm not in any way related to that person. Okay.
Speaker 1:We invented the term cat, basically. Oh, I didn't know that.
Speaker 2:Okay, well, maybe I should claim I'm related and have more credibility.
Speaker 1:He invented the term computer-aided design in the 1960s and his name is Doug Ross. Mit guy permanently imprinted his name in the industry.
Speaker 2:Wow. Well, I can aspire to that kind of goal. I mean, there are a lot of Rosses in the world. At this point I'd definitely rather that metaphor make it smart rather than me personally. But so, yeah, I am not related to Doug Ross my background so I'm a mathematician by training. But it may be worth saying that I never set out to be a mathematician. This is not what I thought I would be doing. If you had told me this is the way things would work out when I was a kid, I would be laughing.
Speaker 2:I really wanted to be an artist and math was just not on my radar. I'm a very, very visual thinker and so I got into math through art. I was really fascinated by tilings and patterns, these two-dimensional patterns and kind of geometry stuff. And then that interest carried me through first an undergraduate and the master's and then finally the PhD. But I've always been very fascinated by the interaction between geometry and applications. And it's a little unusual because I'm more interested in pure math and applying ideas from pure math to problems in the real world. This is in contrast to the field of applied math, which always begins with a question coming from another field of science usually, and there's a whole set of mathematical techniques that fall under the applied math, but that's not really been my interest. I've been interested in theory of geometry and combinatorics and things like this that nevertheless have applications to especially material science. So my PhD was on, really, the structure of zeolite materials that have a very interesting molecular structure and this has led to an interest in lattice structures and metamaterials, which has, you know, inspired a lot of my work in 3D printing.
Speaker 2:But so all that to say mathematician but kind of like an uneasy relationship with mathematics in certain ways, and very interested in art, in design, and so, following, you know, the whole academic path, I met the person who's now my co-founder. This is Daniel Hamilton, who was working. He had started his own consulting business applying mathematics to architecture basically, and so, given my interest, you know, we clicked right away and I started working for him and then we worked together for quite some time and eventually kind of became like a partnership in that business. And then eventually, you know, we had a small team and one of those team members, tom Roslinski, joined the three, the two of us and the three people started Metafold.
Speaker 2:So yeah, and Metafold I guess I can say was inspired by the consulting work we were doing. So in this consultancy, like who is it who is hiring professional mathematicians to solve problems? So we mentioned architecture. But we ended up working, in fact, in sportswear for a company that was trying to do something really ambitious using 3D printing. And it was that context like they're literally hiring mathematicians because they have problems they can't solve about 3D geometry. What are the like, what's going wrong here and what's not functioning about their existing software for CAD, for designing and understanding those shapes. So that was the kind of the motivation for Metafold.
Speaker 1:And now it's been five years and I think I've seen you guys grew at least three years now. I've known you three years ago maybe. What are some of the aha moments for you? Because I know you're very open to new ideas and ready to pivot whenever, um, it's necessary. Like, what are some of the aha moments for you throughout this long journey so far?
Speaker 2:yeah, I mean there have been so many. Jenny, it's like you, you have to be open to that learning, otherwise you will just not survive. So I did mention this one of getting out of 3D printing exclusively. We still support customers in 3D printing and we love those applications, but we're more broad than that.
Speaker 2:I think you know, about a year ago we had a major moment where we really understood our technology as a shape analysis tool. So our technology is not about generating new shapes, and we had been trying to generate shapes for four years and, by the way, we created lots of amazing shapes and helped people design lots of great stuff. But ultimately I think we started to realize that the true gap in the market wasn't in that creation of new things. It was in the understanding of what those things do for a business, you know. Are they going to perform? What is the price? Are they similar to other things, like what is going to be the manufacturing yield based on tolerances? These are business critical questions that come from understanding shape, and so that's a business insight. At the same time, I think we kind of finally accepted that our technology, its true strength, was totally aligned with that. So its strength was not in necessarily design, although we can actually do like tons of cool design stuff but it's real strength is, in this analysis, capability. So that, I think, was kind of the key aha.
Speaker 2:And then, if I can say one more, I mean I could like fill the next half hour with this, but the next one that I would say is on the API point.
Speaker 2:I want to ask you that actually, yes, yeah, yeah, so we did have, and some people will know that we had this app in the market and the app was for designing shapes and it was very cool.
Speaker 2:Like I still use it actually it's quietly on the side but what we came to realize is that it's very hard to build an app that is going to satisfy the needs of all the people working in 3D.
Speaker 2:Uh, and even if we're just taking the people working in 3D printing, those people are in aerospace, they're in footwear, they're in medical, they're in industrial flow processing like they're all over the place, and there's no one size fits all approach that we found anyway that would do this. And so what we realized is that, by providing the API, we were able to provide much more targeted solutions for our customers, and so this also comes along with a change in business model, that we engage very deeply with our customers. We don't have that many customers and we have bigger engagements with a smaller number of them and we make technology that absolutely fits what they need and moves the needle for their business. So it's not about trying to do that a little bit for everyone, but rather get really focused with the API so that we can deliver a very bespoke and targeted solution.
Speaker 1:Do people still have access to the original version, the cloud version, because I still want to play around with it, if that's possible.
Speaker 2:We have secret access, so, jenny, maybe we can arrange something for you.
Speaker 1:Another question I have. From the conversation about APIs, you know, I just discovered there is a small industry, probably also hidden from the public, that focus on geometric kernels. The Russian government apparently owns a geometric kernel that may or may not be accessible to people. What do you think of that industry and how are you guys related to that? Is it?
Speaker 2:the same thing or different.
Speaker 1:Yeah, such a fascinating topic and when you look, this is on Wikipedia there's a list of geometry kernels.
Speaker 2:This is how I learned about it. Yeah, yeah, it is a short list. A's a list of geometry kernels. This is how I learned about it. Yeah, yeah, it is a short list, a very short list. Yeah, once you kind of strike off some of the more niche kernels, it's a very short list indeed.
Speaker 2:So when you think about that, it's remarkable that there's maybe a handful like maybe five kernels that drive all of the CAD software, all the CAE software that drive all of the CAD software, all the CAE software, like everything to do with 3D geometry is basing its form on one of those underlying kernels. So very, very interesting to think about that as a business and also to think about that as how it has shaped the things that we make in the world. Like our tools shape us, so the tools we have access to shape what we we make in the world. Like our tools shape us, so the tools we have access to shape what we've made in the world, and those kernels are underpinning all of it. So, to answer the second part of your question, yeah, metafold, we have effectively built our own geometry kernel, so it's an implicit geometry kernel. We are not dependent on any of these other geometry kernels, but we do interface with them because that's an essential part of um.
Speaker 1:You know, we need to be able to speak the same language and not every one of them actually license their kernels out either, and but on the other hand, there are probably some geometric kernels out there that belongs to some company that we don't really have access to or know about.
Speaker 2:Yeah, this could be. I don't know the story, by the way, with the Russian geometry kernel.
Speaker 1:It's sort of like linked in folklore for me.
Speaker 1:Yeah, it's fascinating. I don't trust everything that's on Wikipedia, but I think the general information on it is accurate mostly, but some of the stuff that's like totally out there, um, you require some deep digging, uh, and also true understanding of the space, which I do not have. Um, all right. So, um, in terms of 3d printing I know you're not entirely focusing on it I would say probably it's accurate to say 3d printing industry benefits more from metafold. The metafold benefits from 3d printing. Is that right?
Speaker 2:maybe. I mean, I think it's just that as a business we couldn't sustain ourselves in that industry, but we're still very keenly interested and happy to support people.
Speaker 1:Yeah well, the reason I mentioned is that I think I want people in the 3d, the 3D printing industry to understand some of the unique benefit of shape analysis, because they probably have a certain type of geometries that would benefit more than the mesh-based traditionally. Totally yeah. So do you want to unpack a little bit of how your geometric analysis would benefit 3D printing?
Speaker 2:Yeah for sure. So 3D printing I mean, the reason we even started in 3D printing is because the geometric constraints are so open here, right? So people probably haven't heard you have free complexity. Well, this is sort of true in the sense that you can probably print just about anything, but can you actually represent it in CAD? And then, more relevant to what Metafold does, can you understand it before you manufacture it? And so I think, yeah.
Speaker 2:So maybe this is the right time to kind of get into lattices and metamaterials. Yeah, this is the and this will be known to many of your listeners but one of the cool things about 3D printing is that we can create lattices and metamaterials. So metamaterials are a way of getting new and different material behavior just using one material. New and different material behavior just using one material. So your three-dipinter is spitting something out and then, by controlling the geometry and the way that it packs together, we can get really interesting and new material behaviors out of that. I honestly believe this is such a huge area of potential, but we're kind of somehow not ready for it. But I'm excited for the day when this can be more broadly embraced.
Speaker 2:But coming back to implicit geometry and metafold one of the great things about implicit geometry and signed distance fields and signed distance functions is that we can represent very complex structures very efficiently Because remember, it's just a function. So when you have a function, it's very different from having a surface representation, like a triangle mesh, where you have to record the details of each and every triangle that is, mapping that very complex surface. So people are familiar with the headaches of large file sizes and processing time and all that stuff, and this is where implicit geometry really shines. You can both represent that geometry but, relevant to Metafold, we can also simulate it and we simulate it in the same representation. So there's no need to take your implicit representation and tetrahedralize it and put it into a solver. We can do this all in one go. So so Metafold and implicit geometry more generally is a great way to understand the kinds of shapes that you want to produce with 3D printing.
Speaker 1:Yeah, um, you said that meta material is too early of a project for you guys to take on. Kind of now, I kind of understand why you're pulling out of the design creation part of the CAD process or the design process, because maybe we just have too many ideas.
Speaker 2:Yeah, I mean, there's still space for this, for sure, and there's some good players already doing this kind of work, I think and we continue, by the way, to support people who are using metamaterials and 3D printing space for this, for sure, and there's some good players already doing this kind of work um, I think and we continue, by the way, to support people who are using metamaterials and 3d printing like it's, it's still an active area.
Speaker 1:It's just that it's not like a market yet and is there any significant difference in computing power between the mesh basedbased representation and the signed distance field representation, these two different ways?
Speaker 2:Well, yeah, I mean it's very compact. The signed distance field representation is very compact, so you can end up with a very concise description of shape that doesn't need like these huge file sizes. You know there's still. There's always a trade-off between accuracy and time. You know, like with sign distance field, in theory they're as precise as you want them to be, but in practice you have to choose a representation to evaluate the function at. So that choice of representation will depend on compute, the function at. So that choice of representation will depend on compute. It's very fast, but as you get more and more detailed, as you get more and you want to do things bigger and bigger, you will pay a certain time cost. But when comparing to meshes, it's very performant, is there?
Speaker 1:such a day in the future is possible to figure out what part of a component use SDF and what part use mesh as a combined product?
Speaker 2:I think this is a great question and I think there are some initiatives in this direction. It's not what we've chosen to focus on, but I think that is the right question to ask, because there are certain structures that are very well represented by conventional, conventional CAD representations and there's some that are better in implicit and so how do you bring them together to find this kind of like happy, happy combination really?
Speaker 1:and maybe also large amount of existing pre-existing mesh that's out there that can use your software to analysis. Now coming to the more, perhaps a little bit futuristic and exciting topic, which is everybody is talking about AI and machine learning these days, even though these topics really existed for 40, 50 years already and because there's money pouring into these names. Can you talk about how you see this kind of wave of excitement behind AI, ml and relevant to Metafold?
Speaker 2:Yeah, it's a good topic. I mean maybe like the first thing to say which I find people are often confused about but Metafold is not an ai platform. Right, we are just not. It's not based on large quantity of data. We're not training models. We're based on that.
Speaker 2:So that's the first thing to understand, but I think, building on that, um, what's interesting is is coming back to again, like our reason for being. We're converting geometry into information and ultimately, that information should be the input to any machine learning models. So we certainly work with customers who are building machine learning models I was talking about the costing models, for example. This is an ML technique and we convert three-dimensional data into a feature vector, something like literally a vector of numbers that a machine learning algorithm can ingest very easily. So the bigger point here, and something I've been thinking about for a really long time, is that AI and 3D do not go together very well, and I even claim and people will, I think, probably push back in some ways but AI for 3D is not there, it's not happening yet and it might be indeed a very long time for it to happen. There's a few reasons for this, but if you think about the LLMs, for instance, the volume of data and the simplicity of that data really is tremendous. We don't have anything like that for 3D.
Speaker 2:And not only that the data we do have tends to be proprietary. There are very few, a very limited amount of public 3D data. So this is one of only one of the challenges but, like the other challenge is that humans, we understand the 3D world so naturally and so easily. And think about the last time you interacted with an LLM and it made a comical error that AI does not understand anything. So thinking about how you would even extrapolate that level of intelligence quote, unquote intelligence to 3D when we don't have enough data and it's just a more complex space, it seems very, very difficult. So I mean, there are subsets of certain applications and subsets and very targeted problems where ML can be very impactful with starting with 3D data, but it's a very small minority of cases.
Speaker 1:I would say it sounds like Metafo basically generated raw data for some of the data projects of starting with 3D data, but it's like a very small minority of cases. I would say it sounds like Metafo basically generated raw data for some of the data projects of various companies, but you do not own those data, unfortunately, so you can't actually create your internal model for any kind of future applications, unfortunately.
Speaker 2:Yeah, I mean for what it's worth. We are working on this. We are actually collecting a data set. One of our current projects internal R&D projects is collecting a data set so that we can build our own models, just so that we have. Yeah, as you say, everything we've done is always on proprietary data, so if we own our own data, then we can just talk about the results a little bit more clearly. But it's very limited and it's hard to find the data.
Speaker 2:It's really surprising maybe not because this is a really new field is that we don't have any kind of open source for the there's a little bit and there's some great data sets that have been created by the academic community and we love them and we use them, but it's sort of not. It's not enough if you think about again, the quantity of data that's feeding these very large models that are having this big transformative impact on everything. All this AI stuff. We think about it all the time. It's something we, as a technology company, we have to keep our eye on, but I kind of always come back to the same point, which is that there's still a strong need for very fundamental mathematical representation and results on 3D geometry. That will be the first step toward any kind of meaningful AI.
Speaker 1:Yeah, it's like building the foundation of future AI. Yes, exactly, and really interestingly, I think any kind of technologies that aim to interact with external world, three-dimensional world, will have the same problem. I read a really good article which I can share in. The link is about how robotics, the field of robotics, especially the ones that we're envisioning, which is autonomous, intelligent reasoning robotics those are not going to happen either for any time. So I think autonomous vehicle is probably going to be the first one, and the major reason for that is creating a data set that's useful for this, and the cost is already enormous billions of dollars, decades of efforts into it and it just simply moved a needle so far.
Speaker 1:And I would assume, like us, our field 3D printing or 3D geometry, this kind of interfacing with external world. We have the same data challenge and software challenge, essentially to really envision that autonomous, automated process. So, yeah, well, fascinating world. I hope the process is accelerating instead of a linear progression, yes, but rather an exponential field. But the good news is to build up that foundation, metafold has tons of business yeah, well, yes, I mean, I sure, I sure hope so so, um, okay, so we're reaching the end of the podcast.
Speaker 1:I have a couple of fun questions for you, okay, um, what are some of the surprising moments, or many, oh oh, my god, there's uh, there's been a lot of surprises.
Speaker 2:I mean. Um, gosh, where to even begin with this? I mean, I think I think honestly, I've kind of already said the biggest surprise, which was understanding our technology for analysis over design, because we weren't I don't know somehow this one was not on our radar, like and it's all in kind of the framing too, like our technology didn't change from having that thought, you know, before and after, but our understanding of it completely changed. So this was a big surprise. Um, there's surprising moments, I don't know. There's more kind of comical things, like, in fact, when we, when we began, we actually built a 3D printer and that's how we got started as a company.
Speaker 2:We were building this really cool 3D printer and it was so great and we were talking to investors and we had this one set of investors who was very interested and we were building the software and the hardware at the same time and they were very interested and they said you know, I think your strength is in the software. Like you've done something very interesting in software. We're not as interested in the hardware because there's like a lot of 3D printers on the market and me and my two co-founders went away and you know like our hearts were in this printer. You know, you get so attached to these ideas and so we had this big soul searching session and then we're finally like, okay, we'll just do the software. So we kind of got our act together on the software. We went back to the same investors and said like, hey, we did the thing where you're pivoted, we're just doing software now. And they were like, oh, this is so great, this is so great. We, um, we don't invest in software.
Speaker 2:So I mean and there's been a lot of a lot of things like this, you know, just um yeah, it's it's just a journey and and I think all the, all the little, the big and small learnings have have kind of been surprising along the way you will think that investor will tell you before. I know, I know, yeah, it felt like, okay, this is going to unlock this next stage, but anyway, we were. We were pretty nice.
Speaker 1:So many moons away. I remember you joined one of our virtual events and I think I don't know if it's because our personal conversation or maybe the events you were talking to something about life, work, balance and you're probably one of the few founders actually talk about that. Not to mention, you're also one of the few female founders who also uses masks extensively. So many unique features. So how do you balance as a founder?
Speaker 2:I guess the first thing to say is I don't know if I really believe in this idea of balance.
Speaker 1:Like if I'm honest about it so it's.
Speaker 2:it's not a balance per se, um, I think it's more that I've I've learned enough about myself to set boundaries and to know what I need to feel like a human being and uh, so you know concretely this work we do. It's so you know, it's in our minds, right, it's in our minds. It's not embodied in any way. It's, yeah, it's easy to kind of get stuck in that world, and so I really think a lot about, just like, movement I'm an absolutely obsessive exerciser, this keeps me sane but also things like handwriting, you know, getting off the computer and and handwriting, um, so the kinesthetic action of, of writing on a piece of paper with a pen, um, and similarly, drawing.
Speaker 2:So I'm a big believer that writing is thinking and drawing is seeing, and so when I want to think, I write and when I want to see which is well, I always want to see, but I just I like to draw and I think that's an important practice as well for me. Kind of quote, unquote balance is like, well, drawing boundaries around my time and then making sure I have time to like be a human, a physical human, in the world.
Speaker 1:Yeah, I have to say you're really a unique person as a combination of artists but also mathematician, which basically is a concrete set of logics, are actually quite inflexible. In a way. You have to say one plus one is equal to two is absolute truth and several steps of logic going towards a certain outcome. So it's, it's fascinating to see both features in one. I have to say my, uh, my brain is malfunctioning, short circuiting at the moment.
Speaker 2:Well, no, but I, I but I, I mean I might push back on that a little bit, because one of the I'm actually terrible at arithmetic, like terrible. Um, that's not the way my brain works, but I think, uh, for mathematics, I think something people miss about mathematics is that it's very creative, very creative. So this is really like the linking theme between between what I do in mathematically or artistically or whatever like, and and also what I do as a business person, like I'm. I'm interested in facilitating creativity in our customers to allow them to achieve these these, uh, these creative engineering outcomes basically. But I think that is for me what I've come to as the linking theme between mathematics and art.
Speaker 1:Yeah, I mean. Some people said that mathematics is nature's language.
Speaker 2:It's universal.
Speaker 1:So well, just for fun, what do you read or observe or listen to to get inspirations? Any kind of external media source that our listener can also try to go after?
Speaker 2:Yeah, I mean I try to read broadly, like it's easy to get into, like the way our media landscape is, it's easy to kind of get into this um kind of place where you're just receiving content that's kind of like tuned to you. So try and get out of that a little bit and I read like more. Let's say, in the past five years, become more and more interested in, in reading long-form books, like just actually books, not long books but just books, and the reason for this is that, like so much of the information we receive is very short, like very short form, and actual ideas, like real ideas, are messier and complex and need maybe a book-like thing to really get into. So I try to read real books. I read them very slowly, so I'll just put it like this I haven't read that many books, but I can maybe give one recommendation of a book I read recently which again is like unrelated to anything else, but it's called Underland.
Speaker 2:It's a nonfiction by this author, robert McFarlane, and he's a nature writer but also kind of a philosopher and a poet. So it's beautiful. It reminds you of why the English language is, uh, is interesting, it's very rich. He like expands my vocabulary when I read him. But Underland is all about our relationship with everything under the surface of the earth. So it's about I think the subtitle of the book is called A Journey Into Deep Time and it puts kind of so much perspective on our everyday busyness when you think about, like, how long rocks have been there, for instance. But it goes much, much deeper than this. He looks at all kinds of underground exploration and the idea of burial and just our relationship with the surface of the earth. It is fascinating and range and beautiful and uh, and that's the kind of thing I like to read is something that's like very different from from my, my, my day to day and uh, and provides like a different context.
Speaker 1:And a bigger view of the world and life Way bigger. Yes, exactly, okay, and I promise you. This is a one last question Do you have any advice for our young listeners?
Speaker 2:Young listeners. Well, too many.
Speaker 1:I don't know from from, uh, I think well, you know, right now, I have to say, we are in this wave, or somebody called bubble of ai, and a lot of new graduates actually in computer science and mathematics can find jobs.
Speaker 2:So this is the rumor, I don't know. Yeah, so maybe you have our data, jenny. I don't know.
Speaker 1:I read something in Wall Street Journal just recently saying that the lower tier computer science engineers- hiring is decreasing.
Speaker 2:So yeah, yeah, so.
Speaker 1:I believe this.
Speaker 2:However, I would draw a pretty firm distinction between that and getting a mathematics education. So I think I guess my advice would be study mathematics. Yeah, not necessarily because you're going to become a mathematician, but rather that it's a. It's a discipline for how to think and how to reason through arguments, and it teaches you a certain way of of approaching the world that I think is focused on searching out truth, which is something I really care about, and and you know asking questions like asking really fundamental questions, that, and you know asking questions like asking really fundamental questions that can yield insights. And so, yeah, study mathematics, become a problem solver, and these skills are indispensable, no matter where you go.
Speaker 1:Absolutely. I love that advice and, honestly, the only class I remember thus far is my math teacher, is my mathematics classes In a good way, johnny, in a good way, and not just learning how to do things, but how to solve problems in many different ways creatively. Thank you so much for joining us today, Lisa. I hope to see you soon again.
Speaker 2:Yeah, it's been my pleasure. Thanks for the invite, jenny. It's been a great conversation. See you soon again. Yeah, it's been my pleasure. Thanks for the invite.
Speaker 1:Jenny, it's been a great conversation. See you next time. This podcast is for educational and informational purposes only. The views expressed do not constitute medical or financial advice. The technologies and procedures discussed may not be commercially available or suitable for every case. Always consult with a licensed professional.
