Podcast: Xometry's Oscar Lovera & How AI Is Rebuilding Manufacturing

In this episode, we speak with Oscar Lovera, Chief Operating Officer, International at Xometry, about how AI and digital marketplaces are transforming custom manufacturing. Drawing on experience spanning factory operations, management consulting, e-commerce, and manufacturing technology, Oscar shares how Xometry is making manufacturing as seamless as booking travel online while connecting engineers with manufacturers around the world. 

Episode Notes

Key takeaways:

• Manufacturing is becoming faster by bringing design, quoting, manufacturability feedback, and sourcing into one connected workflow. 
• AI should augment engineers, automating repetitive tasks while leaving complex decisions and relationship-building to people. 
• The biggest bottlenecks in manufacturing are often information flow and connectivity, not the manufacturing process itself. 
• Great manufacturing partners don’t always have the newest technology. Strong processes, culture, and operational excellence matter more than flashy equipment. 
• Global manufacturing networks are becoming more resilient through flexibility, allowing companies to adapt quickly to changing supply chain conditions. 
• The best engineering careers are built through intentional career choices, strong mentors, and continuous learning, not by following a predetermined path.

(00:00) Introduction to Oscar Lovera and Xometry

(04:50) The Evolution of Manufacturing and Xometry’s Role

(07:53) Challenges in Traditional Manufacturing

(10:42) Technological Innovations in Manufacturing

(13:47) Supply Chain Dynamics and Xometry’s Approach

(16:39) Human vs. AI in Manufacturing

(19:39) Benefits for Partners in the Xometry Ecosystem

(22:39) The Future of Manufacturing and Xometry’s Vision

(33:29) Evolution of Maintenance and Technology

(36:20) Technology Adoption in Manufacturing

(39:49) The Importance of Fundamentals in Manufacturing

(48:37) Hot Takes on Local Sourcing and Data

(53:32) Career Advice for Future Engineers

Farbod: Hello friends,  welcome back to The Next Byte. Today we're joined by Oscar Lobera, the Chief Operating Officer at Xometry. If you don't already know, Xometry is one of the world's largest AI-powered manufacturing marketplaces. Now, Oscar's career has taken him across nearly every corner of manufacturing. He began as a mechanical engineer, spent a decade running factories at Procter & Gamble, and advised global manufacturers as an operations consultant at McKinsey. Led teams at Wayfair, the furniture marketplace, and now oversees global operations at Xometry, where he's helping transform how custom parts are sourced and manufactured. In this episode, we're exploring how manufacturing is becoming as seamless as booking a flight online. Oscar explains how AI is changing the coding process, why instant manufacturability feedback is critical for engineers,  and how digital marketplaces are connecting thousands of manufacturers with customers around the world. We also dive into the future of engineering workflows, where design, manufacturing analysis, coding, and production become a single continuous process instead of a series of disconnected steps. Along the way, Oscar shares his perspective on why trust still matters in an AI-driven world, how manufacturers should think about global supply chains, and the career lessons that took him from the factory floor to leading one of the fastest-growing companies in manufacturing technology. 

So whether you're designing your next product, running a machine shop, or simply curious about where manufacturing is headed, you're going to really like this conversation. Let's dive into it.  

Daniel: What's up, friends? This is the Next Byte podcast, where one gentleman and one scholar explore the secret sauce behind cool tech and make it easy to understand.  Oscar, thank you for coming on the podcast. We're really excited to have this conversation. We've been looking forward to it for a long time. Just for our listeners, would you mind giving us a quick introduction on yourself, your background, what you do today, and a little bit about how you got there too?  

Oscar: Absolutely. Thank you very much for having me. So let me tell you a bit about me.  I am the CEO  at Geometry,  which we will talk about a bit more extensively throughout the session.  Broadly, are an AI-enabled manufacturing marketplace. But I will tell you more because it's not as simple as that. There's much more to that that I love to explore with you guys.  A bit about my background. So I grew up in the Canary Islands in Spain. So sea and sand, a beautiful place to grow up. For those of you who might not be based in Europe and might be in the US, it's like the Hawaii of Europe. It's a volcanic island in the middle of the ocean. Beautiful place.  But  I apparently had enough of that. And I moved to the UK  to study mechanical engineering. I lived and  I studied at the University of Manchester and did mechanical engineering, focusing on industrial topics. And then I spent the first 10 years of my career running factories for Procter & Gamble, which was an amazing experience to learn how to operate manufacturing facilities. I started as a process engineer,  basically looking after the chemical plant that disposes waste of the factory. And then 10 years later, I was running the factory itself. So it was a fascinating journey, a great place. After 10 years, I thought either I'm going to do this for 40 or  I want to do something a bit different. So then I became a management consultant, and I joined McKinsey and Company as an operations consultant. Worked all over the place- fantastic place, fascinating projects, fascinating people, fascinating clients, but terrible travel. But  I spent seven years basically working in operations strategy, digital strategy,  manufacturing, both from a strategic perspective, but also execution.

That was very, very interesting. And then after seven years of doing that, I thought, I'm going to try something else. And I joined Wayfair, which is an e-commerce platform for furniture and furnishings. And I worked for seven years there. Basically, I did kind of everything: a lot of service work, customer service, partner services, and basically getting into the depths of what it takes to win in e-commerce. After seven years of doing that, you can see the seven-to-10-year pattern seems to be repeating itself.  I decided that I needed something different. And then if you combine manufacturing, strategic thinking,  and e-commerce,  Xometry was the perfect idea.  And I joined about seven months ago as the global COO. 

Daniel: I love the segue there of learning manufacturing principles, learning strategic thinking, and then being at the cutting edge on e-commerce. We've kind of beat around the bush, but it would be helpful, I think, for us.  We're both Xometry customers and big fans of it, but there are a lot of people on the podcast who may not be familiar with it. It would be helpful if you could give a brief overview of what Xometry does, specifically at the intersection of manufacturing, strategic thinking, and e-commerce through the lens of your career and what you're working on today.

Oscar: Absolutely. So let me give a bit of context, a bit about manufacturing itself. So, I mean, you know already, but a lot of traditional manufacturing still feels like what booking travel was like 25 years ago, right? You have to go to different places. You're going to have to speak to somebody. You're going to have to wait for them to come back. You hope that it's going to work. You show up with a bunch of paperwork at some point. And that's the way that manufacturing works. It'd be like travel. Now I go on my phone, I click a button, it's booked within 30 seconds, which, when you think about it, is insane. I used to go to travel agents back in the day. Now it's like, I just go to whatever website, and I can book directly with travel. So what we're trying to do is we're building that digital operating system for the physical world. So the equivalent is we take complex  3D drawings. We pull them through AI, we identify whether they're manufacturable, we provide an instantaneous quote for those who want to manufacture it with a price and with a lead time, and then we make the match between what you're looking for and what a partner can do. We provide a match, and we give you the information there and then. Whether it's CNC, 3D printing, et cetera, we offer it to you, and we commit, and then we deliver on the other side. So I think it's...It's amazing. It's an amazing experience compared to what it was in the past, which is procurement of like hardcore. I need an RFQ. Need to reach out to people, get the price. It literally is like instantaneously checking it, calculating the cost, routing to the perfect machine out of thousands of partners that we have. So you can actually get what you're looking for. So basically we're turning custom manufacturing from being a chore to being enabled in an instantaneous way, which is fascinating, right? So effectively, Xometry has been leading the way in this ecosystem, and hopefully we continue to do so. 

Farbod: Well, it's interesting, Oscar. I would say Daniel and I are a little exposed, not exposed, but rather spoiled because we were exposed to Xometry right out of the gate. I graduated in 2019, he graduated in 2020, and the service existed right off the bat. So even for my senior design, we use Xometry to get parts instead of going through the normal flow of like, getting an RFQ from a bunch of local shops and seeing who has the lead time that can actually make that possible.  So we already benefited from this, you know, at this point, gosh, seven years ago. But what that does make me curious about is, you know, like you said, in the travel industry, a lot has progressed where you can do so many things in 30 seconds. How many of the pain points that existed some time ago still exist today in manufacturing that you still feel? 

Oscar: That's a really good question because I think manufacturing has been, there's an evolution happening.  Generally, when it comes to asset-heavy industries,  it takes you take a bit longer to adopt. The reason why the travel analogy doesn't work is generally it's service industries.  Everything can be automated. When it comes to manufacturing, you have a physical asset that you need to interact. I think what we're trying, what the hardest thing for an engineer that uses Xometry is how is it that what I'm doing, how much is it gonna cost and how long is it gonna take? And how do I get the confidence it's gonna be here quick? Why? Because traditional manufacturing, it's like you, the iterative process used to take weeks and months. That's not good enough anymore, particularly for prototyping, right? What you wanna do is like, eh, I want this now; I can test it. Is it going to work? Maybe, maybe not. But if I can get it now, I can test it and whatever it is, I can refine my drawings. If you look at, I mean, the best example to compare it to people might relate to, if you compare about how SpaceX designs rockets versus Blue Origin, that's the polar opposite of each other. SpaceX believes in quick iteration. We're going to break a bunch of them and then afterwards we'll figure out the best approach.  Blue Origin is thinking much more on like,

Okay, how do I actually build it perfectly, perfectly designed, perfectly modeled to get it away, right? And I think, which is what NASA used to do for it, right? I think we're trying to combine the best of both worlds and ensure that the engineers themselves, at every point, have the choice of making those tests, have the choice of being able to integrate quickly, right? And just reducing the time scale, reducing the lead times, improving the quality, and enabling that cycle to be much quicker.

I mean, the other point is generally you're an engineer for a reason. If you don't want to be chasing down partners, you don't want to worry whether the thing is going to work. Want it, whatever you ordered, you want it there. So that secondary point is something that slowly but surely we're getting much closer to enabling, but it takes a while, right? Because on the other side, partners are also trying to figure out how they can automate, how they're going to innovate, how they can integrate themselves much quicker. So we are somewhere in the middle trying to pull these two things together so that at the end it's seamless. 

Daniel: I think I've been spoiled as an engineer. Farbod had mentioned it, right? Having  Xometry at my fingertips from the time I started my career to be able to design something and get an instant quote in literal seconds to understand what the cost of that part is. But one thing that I think that is missed a lot times when people just, they want to go once and just get the quote and not iterate on the design is understanding the, all the complexities that come into play when you have something that feels like a smaller or simple part when I designed it myself, but it creates a lot of risk and a lot of cost in a system because, you know, there's maybe a certain design decisions that I made that I wasn't aware of that makes it really, really challenging to manufacture. I've had some awesome use of the Xometry tool over the last couple of months that I feel like it helps provide some context to me as to like what's manufacturable and what's not. But I would love to hear more, Oscar, and like where you guys are today and also what you're working on in the future to help make it possible for the people who are designing to almost emulate that feedback as if they were talking directly to the manufacturer on this works well, this doesn't, you should do this instead because when I was designing parts for injection molding, you're right. It was frustrating. It took weeks and weeks and weeks of back and forth with my manufacturer to tell me exactly what I needed to do to make it more manufacturable. But at the end of the day, the quoted price that they gave me was like 10 times less than the original price. And it's because I was able to work with them to understand what was within their capabilities. Are there things that you guys are doing technologically to help make that possible for people who are getting an instant quote, but then also to understand the levers that they can pull to make things more manufacturable and more amenable for partners to take onto their plate?

Oscar: Absolutely. I think it's a quintessential question that we get many of our, uh customers and partners to try and resolve. I think there are three layers that we can do with that. Number one is we are processing thousands and thousands and thousands of parts. And actually, we review thousands and thousands of parts manually. We have DFM engineers that review these parts and identify that this corner, this jumper, is going to be a problem; this thing is going to be an issue; that material selection might be poor; these edges are not going to work, et cetera. And we do those things. And in many cases, in most cases, we give the feedback back. If we think it's not manufacturable, we go back to the customer and we say, hey, listen, based on this, we think this is going to be a problem. Maybe you should consider that. Now, the trick is this is not scalable. You cannot do this 1,000 times, but we're developing the insights and the information to identify the patterns, leverage AI, and being able to automate a bunch of these feedback loops, which allows you to iterate much more rapidly. Because you don't want to wait for, I mean, again, you want your design to go out and then immediately try to give you feedback and say, we set a couple of what you should consider. And I think that's first element we're trying to develop is how do we provide that experience? How do we provide that insight that is much more instantaneous when you're developing? So that's one element. Second element is like, we also learn from our partners on what is manufacturable, right? And I think it's how do you then,  once the part goes into manufacturing, how do you extract the learnings from the partners to understand what could be easier, what would be better next time? And I think that second feedback loop, we are starting to work on like, how do we obtain that so we can feed it back into the system? Anything from...It's a quality issue, it's a manufacturing issue, all the way down to like, it was a bit of a pain to do because a five-axis machine is not meant to do that, right? So how do we capture that? So we're working on that. And then the third thing is, and it goes without saying, how do we make the direct connection between a partner and a customer when it's needed, right? Because at the end, engineer to engineer is much better. 

Can you do it easily? It depends, right? If you are a US partner or a US customer, in many cases, you can enable that conversation directly. But what happens when the partner is in China at different time zone? How do you enable that directly? Can you use language models? Can you actually try and summarize the cases and all that kind of stuff? And that's also something we're working on. So ideally,  making that seamless is affecting the perfect match because what's happening at the moment is a bit like saying, hey, I want a room to use a travel analogy, and we are waiting and said, oh, we have no rooms available, and you go back, oh, well, then it should be instantaneous. But you should get the feedback as quickly as that. 

Farbod: Oscar, speaking of that third point, it made me think about kind of the wider conversation of supply chain and availability. I'll confess, supply chain isn't something I ever thought about until the  COVID situation happened. And then the project I was working on kind of became a catastrophe because our supplier was like, we're not giving you any more chips, figure it out yourselves.  And my company had this process of like getting hundreds of engineers basically to overhaul an entire product.  I'm just curious, given the global presence of Xometry, is that something you guys are considering as material availability changes from the eastern hemisphere to the western hemisphere and tariffs and stuff like that? If someone's depending on uh a resource like Xometry, can they, like you're saying, immediately switch over from one provider to another?  

Oscar: It's very important to us, but let me give you a bit of context. Obviously, in the last 30 years, what we have seen is supply chains becoming significantly cheaper and more interconnected. That's a fact, right? That's been happening. However, what's happened is that we've introduced a lot of fragility into the system, meaning that when something in that pipeline doesn't quite work, it breaks, right? And it's like you started through COVID, you're seeing it now through Hormuz, you can see all of those elements. So the question is for us, as Xometry, how do we introduce antifragility, which means that when things go wrong, we are still going to offer you something amazing that nobody else can. And generally, it tends to be three things.  One is, of course, having an amazing supply network that is globally distributed, in the sense of like, you're looking for this component. You don't really care where it is made. You don't really. You might care about where it shouldn't get made, but you don't really care, but you want it and you want to make it. So the first point is like, how do you build a bad network? Second point is how do you make that network successful so that they want to operate in our platform? Because at the end it's a two-sided marketplace. This isn't a, oh, well, I'm going to get your quote and I'm going to go to that partner and I'm going to offer it to this partner. You need to offer it in a marketplace for them to be able to get. So how do you make it attractive for them to be interested in operating in the platform?

And then number three is: how do you make this seamless so that partners also have something they need to work on? They need to come in the morning, they need to look at the jobs, they need to have the information, they need to pick it right. So for us, supply chain itself, luckily because of what we do, the concept of like put things in a boat and they take 12 weeks across the,  that doesn't apply; it needs to fly, it needs to be quite quick transport. But where the partners are, we need to be able to switch it like that.

So suddenly you go and you go like, listen, I am willing to pay X for a component. think that's when tariffs go up the next day. We should be able to pivot that straight away and go into a different partner. So we have a geographic expansion plan. We need to bring those partners up to speed. I mean, it's not a secret, but obviously we source a lot in China. We have a new office in India. We've been operating for the last year. We are growing there, in Vietnam, other places, EMEA, North America, et cetera. So we'll continue to expand that and make it a competitive marketplace. So the better partner is going to actually win. 

Farbod: Question for you, just as you're talking, this is the engineering part of me. That's curious. know, during our day jobs, Daniel and I work a lot with manufacturers and you know, because they run like a specific company or a specific shop, they know what their capacity is and they know like the planning they have coming down the pipeline. Is that a challenge for you guys given that you have so many partners everywhere? And how do you comfortably provide an opportunity to a customer? And do you know potential capacities?  Again, just curious about the engineering side of it all, how it all comes together. 

Oscar: Yeah, it's an interesting debate because I think there are two schools of thought that are extremes. One extreme is if you have enough partners, you will always find available capacity because no, if you have a so that's one school of thought. The other school of thought is more like you need to put capacity with partners and be part of their business. At the end, there's no right or wrong, but basically partners can fluctuate between those two. Those who are much more anecdotal in their contribution, going like the CNC machines are like 89% utilization this month. We need to get the other 11 % out of Xometry, which is, let's go and get some jobs. And others that are like, Xometry is 30 % of what I do, 50%, 100%, right? And this is how we operate. The question is, how do you cater for all of them? Because what's not realistic, unfortunately, is for you to know what everybody's doing at any given time. It's like, oh, he has 80, they have 70, he has 60. It's very hard. You can imagine you have to integrate to 1,000 ERPs. They all have to agree to it. It becomes incredibly complex. The question is, how do you have enough of that spread of partners that are good enough so that you know that your supply is going to be secured? 

Farbod: That makes sense.

Daniel: Kind of looping back to something you said a little bit earlier to my first question, but I wrote it down because it is very interesting to me around having human reviewers for DFM. I think that's one microcosm: a decision to have human reviewers involved in the loop there instead of just delegating it all to technology and delegating it all to AI. In this world where it's sexy and it's awesome to say we're using AI and we're using technology for X, Y, and Z, what are the types of decisions you guys are making? In a world where it's sexy to say I use AI and I use technology to get something done, what are the intentional decisions you guys are making around including humans in specific aspects? And I think DFM is one of those. Are there other um situations where you're including human expertise or would it be helpful even in the DFM case to understand the decision-making that goes behind that on how and when to use AI and how and when to use human expertise, specifically when it comes to manufacturing?

Oscar: Absolutely. That's a fantastic question and is one that we're currently exploring ourselves. So at the end, I'm not going to be the first person that at the end, AI should enable uh and enhance a lot of the work we do. And of course it would take away a bunch of the stuff that necessarily isn't going to be value-added by people. An example: we can get a lot of value out of quick iteration of providing VFM support to our partners. However, in some cases, you just need to speak to an engineer that's going to help you understand how the thing is going to be assembled, how the thing is going to...You want your DFM engineers to do that. What you don't want your DFM engineers to do that is telling you something that you cannot inject mold that. That's impossible. Like, this doesn't make any sense. Or like you've picked the wrong material. That's not necessarily adding a lot of value. So I think for me, the question is like, where can you provide the best customer value or the best partner value? And then focus your people in doing those really, really cool, detailed, granular things that are very, very specific versus using them at scale. Customer service, just to put a very different example, is the same. Like, if you don't need somebody to tell you what the policy for returns is, you can have a bot do that. However, if you need to solve a very complex problem, you want to speak to a person. So maybe next amount of time, that would be completely, completely like indistinguishable. We're not there yet, right? But you still want to speak to somebody who's going to be empathizing with you and someone you're trying to do. So I think what we try to do is balance those two things out and just focus on people and the things that we value the most,  whether it is the firm, customer services, partner services, sourcing, negotiations, whatever. And then other things, honestly, if you look at our partner network, for example, you know what they value the most when we go over and visit, and we spend time with them and we spend time understanding their machinery, right, and I think there is something about at the end, it's a marketplace. They need to trust you and you need to trust them. And ultimately that human connection of like, Hey, this person's got me. I understand. I met Oscar, or I met whomever. It makes a difference. That's going to be very hard to automate. And it also makes a difference when you go and you see the physical location and you see the place, cetera. So I think I've varied; I hope that gives you a very broad answer but focus on what makes the most value to the customer and partners with people in that sense, and then try to automate that scale, the things that you could do at scale.  

Daniel: It reminds me of a quote I read from Charlie Munger, who's one of the greatest investors of all time, but he said that the greatest economic force on earth is trust, right?  It's not innovation. It's not trying to move at a breakneck pace or inject AI at every corner of the business. Trust is the greatest economic force on earth, is what he said. And so I love what your philosophy on this around using AI to help us be faster, but then also strategically investing humans where we need human-to-human contact to build trust. I don't know, that that resonated a lot with me. 

Oscar: And a lot of engineering is based in trust, right? Which is like, you look at somebody else in the angle, like, I think this is going to work. I think we're good. I think we can go. It's very hard to say specifically why, right? But you know, it's like, oh, well, we should go with this. Let's try it and see what happens. And that's a matter of relationship building and sure it can be supported. You don't have to say, hey, I personally reviewed 10,000 drawings and ergo, this is the conclusion. It's like my AI reviewed 10,000 drawings is giving me this feedback. It makes sense. If I were you, I would take these things. So I think the job changes, right?  And you should just empower people to be able to leverage that information that is being given to them more and those insights. I mean, at the end, I was having a daily discussion with my team the other day and it's like,  I, 20 years ago when I was running factories, I was bombarded by data, by numbers. That was my job. This productivity, that reliability, quality. Now I'm being bombarded by choices, which is like, should probably explore this or you should explore that. And my job is to make choices now. Before it was all these data; extract what it matters and what is the insight. Now it's like choice one, choice two, choice three, choice four, why? Most of our data processing now happens through AI and it's telling you, hey, you should consider this. So the jobs change,  right? And I guess it will change for everyone as well. 

Farbod: Oscar, kind of building on that and then extending on. What you were talking about earlier in terms of like, there's two sides of the coin that Xometry's serving. There's the customer and then there's the partner, right? The obvious benefit that comes to mind for the partners is that Xometry is a source of getting reliable leads. But I am curious, especially as you go and you meet with these folks and talk to them, what other benefits have they been reaping from the platform? Like, I imagine me as a uh fresh 18-year-old mechanical engineer,  sending drawings that had literally impossible tolerances to them.  That is one benefit that I'm guessing is really palatable to a manufacturer, that now they don't have to waste their time having those conversations.  Xometry's taking care of it. But I'm curious, what are you hearing from them about the platform and how technology is supporting what used to be a very  friction-full process? 

Oscar: Yeah.  I just came back from China a couple of weeks ago; I spent a bunch of time with many of our partners there. And the value, three things the most. One is access. For many partners, they are a CNC machine shop with like 30 people and 12 machines. They were not getting access to some of the opportunities I would give them access to. They can do it, they can manufacture it, but they don't have the overheads to be able to say, hey, I'm going to go out and fish, like find out what's happening in the US or in US. So the access for them was completely a breakthrough. They were like, wow, all of these opportunities out there, never knew about it, and it took on me. So that's number one. Number two is ease whilst doing that. So they don't have to do marketing; they don't have to do this. It just comes to them, presents to them, and then they make their choices. They say, well, I think that job over there is attractive for us, so I think these 10 jobs or this technology we can do very well, and it allows them to focus on what they do best. Some machine shops, they just want to do parts and parts and parts and parts. Others...They really value complex components because they have a specific machinery use of materials. Others is a hybrid. Others,  like it varies. So what it allows them to do is like not only is it all available, but it's actually relevant to me of what I want to do best. Right. And I think for them that again, before it was like, Oh, this customer has a bunch of lists and they go and they're like, Oh yeah, but this is not what we do. This is not our core business. We specialize on finishing, right? We don't specialize on this kind of thing. It's all there. So they can actually select and choose and pick, et cetera. And then the third thing is support, which is what you're describing, which is like, look, I am familiarized, I'm a China partner, an India partner, a Turkey partner, North American partner; I speak to the engineer there, and we sit down and we do the thing, but that limit should be your market being this big. Now, suddenly, as you describe, it's like, wow, like I know what they want, and I've never spoken to them. It's in front of me. The drawing is good. I understand the requirements. The material selection makes sense. Already, Xometry has highlighted this chamfer, that corner, this thing. And not only that, if you're confused, you click a button, we help you. Say, hey, I'm confused because the 2D drawing shows this, but the 3D drawing shows that. So this material selection, I'm not sure. Go to Xometry, and we'll help you. We'll resolve with the customer and come back. That, for many of them, was a complete breakthrough. That's why partners operate with us, right? That's why they choose to work with Xometry. 

Farbod: No, it makes sense. Especially when you were talking about like the different time zones and the language barriers and stuff like that, instead of, you know, someone in California trying to work with someone, you know, in the Eastern hemisphere, you easily just bridge that gap within one place. 

Oscar: Absolutely. And again, they want to do what they do best. And that is not necessarily customer service, partner services, marketing, sales. They want to manufacture. They want to do it really well. And they're very proud about what they do.  And that's what they want to focus on.  We enable that. 

Daniel: One thing that I think, if you'll humor me and take us down a little bit of uh a rabbit hole here. I'm curious,  looking back earlier in your career, right? Running manufacturing lines, it was at P &G, correct? Yes. Running manufacturing lines at P&G, um, and the experience you got there in, let's say like, is state-of-the-art manufacturing, but...traditional in the sense of not enabled by quoting an instant job getting by things like Xometry, traditional manufacturing where they have long-term production plans and long-term production. How do you see differences between what you did in the beginning of your career and how you see manufacturers operating now? And to what extent are technologies like Xometry enables that?  And I  guess I'm curious in general, the evolution of the manufacturing ecosystem and where you think we are and where you think we might be headed. 

Oscar: Absolutely.  That makes sense. So, I spent my first 10 years working in heavy industry, right? So it's like a heavy traditional machine paper-making kind of industry. So it's like a big mill, and it's got hundreds of people, all that kind of stuff, right? So it's very different from custom manufacturing and CNC machining, injection molding or 3D printing. However, some things are the same, right? So process excellence, what it takes to be a successful manufacturer, the basic principles, people leadership, importance of objectives, key results, performance, like all of those things are the same. Why does that matter? Because when I go to a facility, it doesn't matter where I go; some of the basics are the same, right? And I think that gives you credibility in being able to identify this is a good place, this is going well, there's a problem here, irrespective of the technology. So some things are common. But other things are very different when you do, like high-speed FMCG, like fast-moving goods, consumer manufacturing is very different to what we offer from a CNC machine shop, which is custom manufacturing for specific components. So some things are not quite comparable. However, let me tell you a bit about the evolution of what I used to do compared to now. Examples are, we used to carry all of our maintenance parts and all of our stock in our warehouse, meaning every single component that I have- and this is millions of components; they'll be spare somewhere. Now I can print them, or I can get them manufactured overnight. That was impossible before. Impossible, right? Secondly, we did time-based maintenance. Did like CBM, like, know, condition-based monitoring, all this kind of stuff. And it was completely disconnected from the world. It was completely disconnected. It was just somebody with a probe goes in, sticks something. Now you can connect it. It will go directly into somebody else. They'll review the maintenance procedures. They'll tell you what they need to. You should probably order a component in three months, two months, whatever. 

Back then it was completely impossible. And then the third thing is like you were tied to a vendor. That's it. There were no competitive prices. He was like, you need that? You need a bearing? This is the only bearing you can buy. You need a shaft? This is the only shaft you can buy. And not only that, we used to make them ourselves. It was like a machine shop of some maintenance guy that used to go and, like, mill a shaft with something broke. All of these things are incomprehensible now. Now it's just like I click a button, I upload the drawing. It comes in the next couple of days; that's done, right? So I think it's evolved a lot in the element of like interactivity between things. I basically operated an island. Now, It's a whole continent of interconnected networks that used to not exist in the past. 

Daniel: I appreciate the context there and definitely for Farbod and I inour day jobs, can, we can align with a lot of that and seeing even small shops now having access to lots of technology to help it, you know, make condition-based monitoring automated and stuff like that. It's, it's wonderful to see that not just, uh, being bombarded with data, right? You said you were bombarded with data, being akin to insights, real insights, not just data, but actual insights that help you make decisions; that's a game changer for manufacturing. And looping back to something you said at the very beginning,  you mentioned that there's some hysteresis in the system. It takes a little time for technology adoption to fully take hold in an industry where a lot of the equipment is physical, right?  There are pieces of equipment that will operate for decades and have operated for decades, and they don't need to change anything the way they operate that equipment to continue to make money the way they've been making money. So there's a little bit of hysteresis in the system. I'm curious from your position, because you interact with lots of different partners, you interact with lots of different segments of industry, and then also your customers, in many cases, um you know, sometimes it's an individual, but in many cases, it's a cutting-edge company wanting to get a uh solution for manufacturing one of their parts. How do you see the technology gradient across manufacturing?  Is there a lot of manufacturing shops that are fully adopting technology and there are a lot of shops that are slowly adopting technology? What is the difference between those shops look like in your mind and where do you think, um how long will you think it takes for hysteresis to work its way out of the system completely?

Oscar: It's a good, it's a good question. And partly is where you're describing regarding how quickly does the equipment turn around. And I'll give you an example. If I go to a 3D printing facility, it's a spaceship, right? It's interconnected. You can upload their face. Why? Relatively new technology. And generally, 3D prints tend to have a relatively low lifespan. If I go to a CNC  shop where these machines last 10 years, and then about seven years ago, it feels like they were operating Windows 95. It's like completely two different extremes. So it's much more and you can see the difference. You go to a 3D printing shop and it's like there's a screen and you can see the model in the screen and it highlights things and you can press the things and it optimizes itself. You go to a CNC machine shop and it literally is a keyboard. Like you're pressing the buttons and it feels like something out of the 1980s. So there's an element of like how different this is. Number two is also when the machinery gets produced, how enabled those manufacturers are, those OEMs are. And in many cases, some of the more traditional technologies are taking too long to figure out how do you disintegrate into the wider ecosystems. I'm sure they exist. I'm sure if I go to one of the treasures, they'll show me the CNC machine that directly connects to Xometry, for example, right? And you send a drawing and it automatically uploads it there. And it just gives a- it doesn't happen, but I'm sure there will be. 3D printing, it works that way. Injection molding is getting that way too. So it's kind of evolving. But ultimately, it's an ecosystem question.  And let me just explore this a bit further with you. Do you know who the best partner that I saw on my last visit to India was? Was a shop that was only like four or five CNC machines, but the layout was perfect. It was like a U-shape. Everything flew. They had a computer, which, no word of a lie, the screen, the monitor, was a CRT monitor. It wasn't even like ancient. But basically the orders came in, they printed the order, they gave it to this next person, they did, going to the QA lab and within a day they could do anything. Right? And it's just very lean, super tidy. And they were talking to me and they were like, well, we're going to expand the facility. And I was like, I love it to expand the facility, but don't change this because you are miles away and the technology level was very limited, but the setup itself was just perfect. Right. And I remember walking out going, that's probably the best CNC machine shop I've ever seen. So although technology can enable a bunch of things,  best manufacturing practices, particularly what we do, and ultimately good culture, good systems, good processes- already that's where you see the biggest differentiation. I've also seen some- I'm not going to tell you where, but some 3D printing shops that have the latest, latest, latest technology and they are running it like it's like a 1970s car, right? And it's basically like a spaceship using it as a bus, right? No disrespect to buses, but you don't use spaceships for buses. And you're like, it doesn't matter how enabled you are,  without the right processes, systems and culture, you're not gonna make it succeed. Now, if you combine those two things- technology- and then you are dependable. But honestly, maybe I'm showing my age and I'm showing my experience. Without the fundamentals, technology becomes ancillary. You need to have those fundamentals. 

Farbod: No, I mean, I think that is actually like a perfect testament to what we're seeing today as well, right? Like everyone talks about AI, for example, and AI and any technology amplifies whatever you got. So if you're good, like you're saying, if you have good fundamentals, then it will make you so much better at what you're already doing. But if you have poor fundamentals, then it's just gonna make you be putting out stuff worse at a higher rate. Like, that's what it comes down to. Oscar, one of the things I was gonna ask you about, I can't let go of this travel analogy that you used. That's how good it was.  But, you know, like let's consider traditional manufacturing from 20 to 30 years ago where, you know, there was Xometry or whatever as the starting position. And whatever the perfect solution is, as like the ideal vision, know, the 100 % mark. Two-part question for you.  One, where do you think we're at in that journey? How far along are we? And then, uh the second part is, what can you share with us in terms of like, you personally seeing, as a wishlist item that'll, like, help us get there? Or if you can even share like what's coming down the product pipeline that's gonna make this process even better. 

Oscar: Absolutely. That's a very good question. think ultimately you want to keep, so there's a lead time and there are costs associated with all of these things that occur by nature. Not everything is instantaneous. But what you're trying to do is ensure humans are only there for the critical decisions where they can add the most value. And an example is, for example, let me give you an example actually. We have a recent partnership with Siemens that we announced recently on uh on social media, publicly etc. Why does that partnership make sense? Because the Siemens design suite, right, that we currently have that I currently have if you enable it through Xometry Engine for quoting as you design you can get a quote. So you're like, oh, if I do this, how much would it cost? Oh, okay, let me do that. How much would it cost? Like suddenly you've taken two steps, decide, unquote, and you put it into one. That's true all across the line, all across the line. So an example would be just to use the analogy that you used before, you'd be like, well, actually, why don't we, instead of presenting the jobs, right, that any job that could be available, if the machines that the partners had, were connected to the internet of everything, they could tell us, hey, I can do these things and I think those jobs over there make sense. Uh, I'm going to present those to my human person to say, listen, these jobs actually have capacity; they're little, make sense.  And we could probably feed them in on Wednesday. And actually, these materials, we have them in the ERP. So why don't we do that? So like,  I mean, I'm making an exaggeration, but all of these little decisions that get made, you can somehow bring it all together which reduces the lead time significantly. Because every time you take a step, you create a queue. Every time you create a queue, you inject errors, problems, confusion, lead times, et cetera. And I think it's how do we connect this? So in the future, in an ideal world, a lot of this is automated, unless there's something critically important for uh customers or for partners or for us to get involved in, right? And I think that at the moment is very much enabled still by people. Right? Which is like, hey, we're going to put this in the marketplace and people look at it and go like, oh, I want to be this or this and this and this. And I don't want to do those things, et cetera. I think it's much more we can do in that space.  I mean, specifically what we're doing to enable that is what I mentioned before. We have this partnership to integrate this a bit further up in the design funnel. That will be cool because it actually stops you from going, okay, well, I want to design this. Okay, we're going to, oh no, we have to go back. Suddenly you can do it real time. And then if you're interested, press a button; we'll go and source it for you. That's amazing, right? 

Farbod: Correct me, chime in if I'm wrong, but  is the vision that you're painting here is that what used to be completely discrete, separate processes between design,  DFM, planning, manufacturing, quality coming back, coming and melding closer and closer into one sort of flow eventually? 

Oscar: Correct. Gotcha. Which, if you think about it, that's how it works. That's how travel works. It's like you go to your app, you press up, do you know how many intermediate steps use to be there? You have to go to the engine, the agent will look into the computer, the agent will then speak to the hotel, the hotel will speak to bookings, bookings will tell you where they're from. Now it's just like, beep, and you booked that room. It's the same. Same thing. And it's like, how do you get as close to that as possible for custom manufacturing? 

Farbod: Well, I'm even thinking about like, you know, doing the mechanical design of it all. It used to be that,  like I had an experience where the mechanical design engineer made the part, we had a review, then it went to the DFM feedback and it was poorly made. So then, like everything had to start from scratch again, right? So another internal review before it went for DFM. Now I'm thinking if it's all done in the same platform, not only are you getting the cost and DFM feedback right away, but by the time you lock it in and you're getting approvals from everyone else, it's like everything that needed to be approved is done. And now we're actually making this thing. So, closing that feedback loop is very promising. 

Oscar: For me, I think it's critical because I think the more you want to make this as iterative as possible, closest to the source. Yeah. Right. So you make the choices as early as possible. That makes then a straight line all the way to manufacturing and back. The problem is when you're doing them all throughout the journey, then you have the problem of like, ah, we're stuck here. We need to go back to square one. Like snakes and ladders, you go all the way back. Yeah. That's, and it's just like, how do we make it go straight through? 

Daniel: I love that future. And  I'm going to be biased here and people who listen to the podcast know that this is a bias that I have and it's halfway joking. But I often say that, that engineers are the only people in the world fighting against entropy. You know, entropy is trying to destroy the entire world and engineers are the only people stopping it. It's not entirely true. There's plenty of good people who are trying to stop the world from becoming worse. But I love a future where engineers, we reduce the amount of iterative cycles or reduce the amount of lead time in those iterative cycles so that engineers can work faster and design cool things faster. There's a lot of butterfly effects that will come from that, just from people being able to come to solutions faster. We don't even know all of the impacts that will come from that, except that I know that people who are working hard to solve important problems will be able to do them faster. And that's really important to me. So I love that vision.

Oscar: It is. And I think for me, one of the fascinating discoveries I made here was starting working with manufacturing partners that manufacture drones. And you tend to think about this kind of industry. And I know drones are not planes, but generally, you think about it, it's going to be a design, and you're going to make 1,000 of them, and they're going to be under that. They're iterating every week. Every two weeks, they're going like, OK, we're going to do this, it's going to do that. And then they test it and go like, yeah, let's try again. And then, yeah, let's do this. Or maybe this. And then suddenly, four weeks, new drone technology. And that for me, 25 years ago was incomprehensible. You couldn't do that stuff. Even if you tried, even if you had all the OEMs around you, you couldn't do it. And now there are people doing it every day. And they're like, I'm going to make 15, and I'm going to test them and I'm going to break them. And then I'm going to make another 15. And they do it in days, weeks. I think that's more of the future of engineering. I understand certain things don't work that way, but many things do. So enabling those things is important.

Farbod: That's awesome. Oscar,  one of the things we always ask on the podcast, out of curiosity, especially since you have such a wealth of knowledge from the manufacturing industry, what's a hot take you have about what's going on and whether it's something happening right now, what you think the future should look like, et cetera, et cetera? 

Oscar: So I was thinking  when you mentioned about hot takes, let me give you two. Let me give you two hot takes. Hot takes, right? Okay. So I think that local sourcing is often a false safety net. I people think that proximity equals better. And I don't think that's the case. I think making the perfect match between you and some partner is what makes a difference and is how do you enable that globally? Right. And I think that for me is more important. Oh, no, we need to keep it close because it makes it. I don't think it does. I think you could be anywhere in the world. That is my job is to try and enable that globally, which is like, don't care where it's short. I just want to have an amazing partner. Whether it's here, there, could be local, but it doesn't have to be local. And I think that's a mindset thing that people have and like, yeah, that's true. But you know, it's much better to have them beside you, right? I don't think that's the case. So that'd be the first one. And then  I think the second one is that I think the vast majority of problems that occur in manufacturing, in the point from you actually wanting something and something coming to you, have nothing to do with the process of manufacturing, but it's failure of data and insights. I think it's like very rarely I see somebody that says like, oh yeah, but this is dark. There's always a why behind it. And generally it tends to be that connectivity, which is like, how do you ensure that the needs and the capacities and the capabilities are met perfectly, that perfect match? I think that's the sort, not having that is the source of vast majority of problems. Not the five-year machinery shutdown or it's broken or somebody did something wrong. I think that's a minor part of it. 

Farbod: Speaking of the hot take, if you want to make it hotter,  would you say the lack of data is typically on the customer side or on the provider side? 

Oscar: I think in both. I think we struggle on both ends. And it's not necessarily the customer knows what they want, but the question is: how do you convey what's here into a way that can be processed? Um, and I think the other side on the partner side is also like, how do you get that and convey it in a way that's understandable to you. That's what we're trying to do, right? Which is that how do we provide that connectivity? But obviously more data doesn't mean it's better, but the right data makes it so that that match is good. 

Farbod: And then your first hot, your first hot take about local, that was properly hot. I was going to ask, even with like, you know, tariffs and stuff going on all over the world, it's hitting everyone. Do you still hold that opinion? I think so. I think the question is, how do you- how do you inject antifragility into your system? So when things happen, you're stronger, right? Not necessarily weaker at the moment. Most of our supply chains are robust, meaning that they can take the hits, but anything robust breaks. What you want is something that takes hits and gets stronger because it learns, adapts, it evolves, and allows you to be better. So for me, I think it's, and ultimately at the end where you're looking is if a customer absolutely demands something, of course you're gonna provide it to them. But the question is why are you demanding it? And very rarely it's driven by a specific; it's like, look, I want this cost, I want this material, I want this thing, can you give it to me? I think thinking that that is better done locally is not necessarily the right mindset. I think it's a false safety net. 

Farbod: Well, we asked and you properly delivered not one, but two very hot takes. 

Daniel: I love it. Oscar, one other question that we ask to almost every single guest in the podcast, and I'm going to zoom out and give you some context. Um, our podcast, we say that our mission is to try and find the secret sauce behind a lot of interesting technology and make it not only approachable for people to understand, but also interesting for them to understand. And so that's what I say when I mean the secret sauce. Do you have any secret sauce on you? On your career, on your journey, um, that other future engineers might need as manufacturing becomes more data-driven, is more precise, more connected. Looking at, from you studying mechanical engineering, if you were to start your career again today, what's the secret sauce in ah, architecting a career that gets you to be the COO of international business at Xometry? What are the few ah levers behind that and advice that might be relevant to someone starting their career today?

Oscar: Absolutely. So let me give you three things that I think are universally true. uh Number one is that, number one is nobody cares about your career but yourself. And by that, mean, of course other people care, right? But universally, you need to have a perspective of what you want to do and take away the myth of like, well, I chose to do this. I'm going to do this for the next 40 years. That's not the case, but what do you want to do next? What do you want to explore? It's not your boss's job to tell you what your development plan is. It's yours. So first of all, if you want to have a good career, you need to be consequential and develop those plans yourself. And like I said before, I ended up here not because I was a phenomenal engineer, but because I took a job out of engineering and I started doing consultancy. And then I took a job after that that was e-commerce. And now I'm here. Those were conscious choices. It wasn't momentum. was decisions that I made myself. So, number one is important for you to have a path of what you want to do next. Secondly is, and this is probably something that is underestimated still now, but it's like there are a bunch of good people out there that can mentor and help you. And people tend not to ask enough help. I don't know whether it's a sign of weakness or a sign of like, I feel uncomfortable. I am the person I am because I have phenomenal mentors, but none of them came to me. I had to go to them. I had to go and understand. I can name them now; they're so impactful to me. And I think that's important to understand who are you looking up to and what are you learning from that person? In some cases, how do you solve problems? In other cases, is their technical expertise or how they manage people or how they lead organizations and focus on understanding how they do it and practically help out with that kind of stuff. And then the third thing is like, whatever you think is gonna look like in five years, it's gonna be different.  And I think the world has always been changing. Right. And people tend to think that, oh, we live in a world of it's always been changing and innovating. But what I've experienced in my last 25 years is it's happening quicker.  It's like the cycles are much faster. So being agile on your feet about what's happening, learning new things, exploring new topics and all that kind of stuff. Absolutely critical. was important to me 25 years ago. Now it's critical. It's absolutely pivotal for you to keep up to speed and see how this interconnection works with you, et cetera. Now, I don't think those three things are just for engineers. I think engineers as well. But I think they're fundamental for the success of any career.

Farbod: And you really packed it on with the hot takes and hot advice. 

Oscar: Well, I hope is useful. I hope that your viewers and your listeners appreciate it. 

Farbod: I'm sure they will. 

Daniel: That was great. Thank you very much. Thank you. I think the last thing here on my end before we wrap up is I always want to ask, there topics that you wish we discussed, or are there things that you want to share specifically on what you're working on today, something that's interesting that we haven't shared yet? And it's okay if you say the answer is no. And also, how can folks find you if they want to hear more? Cause I anticipate that folks will want to hear more of Oscar Lovera after this conversation.  

Oscar: I am more than happy. You can find me on LinkedIn. You can go to our website, Xometry.com, or whatever country you're in. You can find me anytime oscar.lovera will be available for anyone who wants to contact me. I love doing these kinds of things. It's very rewarding. And I think for me, the interesting and I'm very excited to meet both of you because what I'm finding is there is a resurgence in the interest of our trade, of our industries, right? And I'll tell you why: when I graduated from mechanical engineering back in 2003, so, 23 years ago now, 24 years ago now, nobody wanted to be an engineer. Everybody was becoming a banker. So out of all the top students of engineering, they all ended up working in banks. Engineering was like, yeah, yeah, why would you do that? It's dirty, it's long hours, it's hard work. Why would you do that? And I think I'm excited to see young engineers that are keen to develop technologies, that are keen to understand more, kind of like getting their hands into this industry, because it's fascinating. And I think what I'm interested in understanding more is how do we get more? What needs to be true?  Because it is certainly, and I'm sure you guys have noticed it, there's not that many. There's not that many that want to do what we do. And it's so rewarding, and it's so kind of like energizing and everything.  And I need them myself. I need more engineers that can work with us. So I love to explore that more. What needs to be true for us to expand the community further than what it is now.

Daniel: Well, we appreciate you participating in it. That's a big part of it is there's a lot of folks, a large contingent of our listenership are folks who are in engineering school now and are figuring out what to do next. So appreciate you sharing a little bit of your story and some advice. think that, that at a minimum, will help contribute a little bit to having some more well-equipped, well-adjusted engineers that are going out to solve real problems in the world. So thank you. 

Oscar: Fantastic. Thank you very much. Appreciate it. Thank you very much for your time. And yeah, I'm happy to do a second episode whenever you want it. 

Farbod: That's what we like to hear.  Thank you very much. Awesome.  

As always, you can find these and other interesting & impactful engineering articles on Wevolver.com.

To learn more about this show, please visit our shows page. By following the page, you will get automatic updates by email when a new show is published. Be sure to give us a follow and review on Apple podcasts, Spotify, and most of your favorite podcast platforms!

--

The Next Byte: We're two engineers on a mission to simplify complex science & technology, making it easy to understand. In each episode of our show, we dive into world-changing tech (such as AI, robotics, 3D printing, IoT, & much more), all while keeping it entertaining & engaging along the way.