NATAN: Lior, hi. How are you? LIOR: Doing well. Good to see you. NATAN: It is good to see you. It's been a little while, you know. So, Eclipse is now how old? LIOR: Almost eight years. NATAN: Wow. I remember we were, like, back probably eight years ago sitting around somewhere probably in Oren's Hummus [SP] or something like that, talking about how venture capitals don't [laughs] understand anything about investing in industrial anything. LIOR: Probably the people that sit next to us thought, who is these people? They must be not from Silicon Valley if they're talking about manufacturing and industrial. NATAN: Yeah. Thanks for coming on Augmented. It's great to have you. You know, this podcast is really about what's happening in all things industrial from many perspectives. And I think bringing people who are focused on investing into startup in the space is always very exciting. We had folks like Dayna Grayson before and a lot of perspective, another, you know, early-comer in the industry. And now, you know, when I think about Eclipse, I'm thinking about companies like Augury, and Instrumental, and Bright Machines, and a bunch of others. So, how do you define what's industrial and fits your portfolio? LIOR: When we started the firm in kind of mid-2015, the rationale was very fundamental based, although I was not sure that I understand what the fundamental investments mean back then. But it was, if you're thinking about the world GDP, you're talking about $100 trillion today, give or take. More than $80 trillion is what I will categorize as the physical industries or industrial industries, and that's things like logistics manufacturing. It's aviation; it's defense; it's agriculture and mining; it's shipping. It's where the economy is actually moving something in the physical world, not in the internet, to generate something. And those industries that carry more than 80% of the world GDP they just did not have the digital transformation yet. And that's kind of was the reason we started the firm. NATAN: I remember you telling me about the genesis of this all the way back to Lab IX at Flextronics. Is that where things got shaped up in your brain to think about this space from a perspective of venture capital? LIOR: Yeah, absolutely. And, you know, I did not plan to be a venture capital. I didn't go to Stanford Business School or necessary was an associate in Sequoia. I built my way up. It's not something I had been taught. NATAN: Well, that's good because I hear they're letting people go nowadays. LIOR: [laughs] Yes, not a joke. But yeah, when I was at Flex, and I was in charge of lab IX, the digital transformation team at Flex, and Mike, back then the CEO, hired me to come and build that team. Flex serving exactly those industrial industries, right? So, they're doing more than a billion dollar in energy, in manufacturing, in automotive, in medical devices, and aerospace and defense. And I had the privilege to meet the CEOs and COOs of some of those Fortune 100. And none of them would want to talk about manufacturing and Flex services. All of them will want to talk about technology and how it's changing their interface. And I had this weird feeling...Because I lived in the Bay Area, everyone is telling me software will lead the road and crypto, and now GenAI. And I'm like, yes, but those industries are actually the ones that running the world. And they did not had a digital transformation yet. And that dissonance is what sparked in my head the reason to start Eclipse. NATAN: Yeah. And operations, manufacturing is so physical, and so, at times, you know, super detached [laughs] from all those big words around cloud infrastructure this, and data like that, and all those, like, mega buzzword trends. Do you remember, like, an industrial aha moment that you're like, 'Oh my God, there's so much value here,' that you kind of met maybe on the floor in Flex or in one of those interactions? And how did that actually translate, for example, say, for a startup you backed? Because I think that's the thing that a lot of people don't ever connect, you know. LIOR: Yeah. We have so many of those stories. I will give one quickly, and then I would love to talk about the largest companies in the world. Because I think, for some reason, I feel there is still, in people's minds, mainly in the worlds of VCs, this notion that if it's not purely on the internet, it cannot be big. And then, you're literally looking on the data and see that this is just not true. And, naturally, Tesla, and SpaceX, and Nvidia, and Amazon all companies that are running physical operation and doing quite [inaudible 04:41] in the public market. But going back to Flex stories, actually, it's touching in Tulip also in some way. We ran a lot of front-end line of SMTs to build their PCBAs. In Flex, out of 150 factories, about 50 of them have those SMT lines, and we never used the data. You had an operator on the line that was tuning the machine, kind of based on what he was...machine does. And we had this idea of, like, hold on, if we can take the data of the SMT, put it in a very, very basic algorithm to find anomaly detection, can we actually improve in 1% the yield of those front-end line? We turn up actually improving in 3% and saved Flex close to $100 million in bottom line every year. It will be 3% operating profit, a lot of money on the top line, about 3 billion. And it was just exciting. And it just made in my head clear that the opportunity to leverage technology in an industry like manufacturing didn't even scratch the surface. And the impact is remarkable. NATAN: Yeah. I have sort of an interesting complement story to share on that. It's just a little anecdote from a conversation with a Fortune 500 CIO because we were, like, making fun of this notion of ROI and perceived ROI because everybody comes in, and the ROI is going to be, blah, blah, blah. And the vendor says this, and, you know, the buyer says whatever. And, like, nobody believes anything anyway [laughs], you know, not the vendor sharing the ROI and not the CIO receiving the ROI. And what was, like, really profound because, like, the conversation I'm having is not about, like, I want to sell them some BI tool or some, I don't know, cloud infrastructure and that kind of thing, it's more, like, you got to enable operations. And so, the guy explains to me, if I can attach the investment in the system to something that is actually happening in operations that often amounts to, like, some lean type of net benefit calculation, you know, reduction of waste, increase of throughput, what have you, then IT people can support OT. You can imagine, like, why we didn't have any collabo...or not, I mean, let's not be so radical, but like, very problematic, say, collaboration between IT and OT for many decades because, like, they're just chasing other things. And to hear that kind of thing from a CIO, it's actually pretty cool that that's the thinking of ROI, as they consider investments in systems for operations. LIOR: No, it's...we spend a lot of time on the engagement between IT and OT, information technology and operating technology. And we think, in the world of industrial, that linkage is so powerful. And historically, essentially, in the last 20 years, all of the tech progress has been done on the IT side. And I think in the next 20 years, it will be mainly on the OT side and then connect between the two. And we should have a much better overall world GDP goal if that happened. NATAN: Yeah, that's actually a good segue for the next thing I wanted to chat about where, you know, we don't have all the data sources to cite, but you know, this cold. And we'll include stuff in the comments after the episode. So, from a very sort of high-level overview, I think the general phenomena that there's a complete discrepancy between the amount of venture capital going into traditional IT B2B type technology landscape, whatever your marketing stack, HR [inaudible 08:10], obviously, now GenAI, whatever, et cetera, cloud. Versus let's invest in industrial and/or with, you know, the hardware component, be it sensors, or real manufacturing tech, or 3D printing technology, whatever that may be. What can you share, like, from your perspective, you know, running a fund that is so focused on that? How do you analyze that phenomena? What does it look like in numbers? And why did it actually happen, and how is it changing? LIOR: So, we started the firm in 2015. And I will tell you, we were alone. Like, nobody would want to hear about those industries. And, you know, we had some big, quite...it was always clear to us that those industries are huge, but we were like, hey, maybe we are too early in this phenomena. We were seeing, like, from 2015 to 2016 to 2017, like, 20%, 25% growth every year investing in industrial because there was other long-term trends that was happening like electrification. Tesla was already doing really well. Like space, SpaceX was doing well. A bunch of the semiconductor companies was doing this, and Amazon was pushing logistics farther and farther. So, it started, but it was not, I would say, step change. But I think we really were seeing going from double-digit growth to 150%, and then 200% and 300% year-over-year growth in investments in that space happened in 2020 when COVID hit. Because I think what happened is you had a ripple effect of the world understanding that there is a massive lack of resiliency in our backbone economy and that we shut down factories. We cannot produce cleaning equipment, and toilet papers, and meat. The supermarket was empty. We couldn't move the products. We couldn't move along the chain. And the government, I think, woke up and said, hold on, we need to make some significant change into the infrastructure that we're living. And it's going to come with a massive funding in this country, a trillion-dollar infrastructure bill, in many other countries, a different type of bill. And from 2020 till today, we're just seeing a hockey stick of investments in the industrial market. Yes, it did help that Tesla arrived to 700-800 billion. Yes, it's helped that Nvidia crossed a trillion. It's helped that Amazon continues to grow. And, you know, Apple is a hardware company. And a lot of those companies that, you know, being a marquise, we usually don't stop and think about Apple as a hardware company, right? We think about it as a consumer company. And it's like, no, no, it's a manufacturing company. And Tesla it's a manufacturing company, and SpaceX it's a manufacturing company. And I'd equate the success of those companies with the geopolitics issue, with the global warming, with the COVID bump, with the war in Ukraine is pushing a massive amount of funding that I think will power the next industrial revolution. NATAN: So, this is fascinating. And, like, we're going to try and share some numbers, like we said before. And, you know, the other thing I see in the numbers often, which I think is pretty accessible, is that it's also about, you know, and what we call frontline operations, about 20% of the global workforce, which is, like, an insane number, right? Like, if you compare it to, you know, the amount of workforce in commercial, so, you know, marketing and sales, development, and stuff like that, which, is, like, somewhere between 3% to 4% or 5% tops, depending on how you count, it's orders of magnitude, right? And so, the impact is not just on the, you know, dollar spend into the infrastructure. It's actually impacting a ton of people globally, across regions, countries, cultures. LIOR: And how much capital is being invested into marketing people versus frontline. NATAN: Right. LIOR: Historically, from the venture community and the tech companies. I had a question for you, slightly different question, but also on the frontline workers. But I will let you finish the thought before I ask the question. NATAN: I mean, the thought is that you know, we often think about the digital transformation and the initiatives and, like, the big bills, and all that kind of stuff, and the output is on, you know, society level infrastructure companies and all that kind of stuff. But there's just, like, a ton of people involved. That's, I guess, the point I was trying to make and give it a little bit of a context with the number. LIOR: No, absolutely. We've been asking ourselves that question and also have been asked that question many times. How do you view the frontline workers influenced by technology? As you know, we do a ton in the world of automation. And people are always coming and saying, "Oh my God, it's going to replace all of these jobs." And I'm like, "Guess what? No people that want to do those jobs." And I think what we want to see is technology like Tulip coming and actually boost productivity so a human can do much more than he could see before, that he can do before. But I'm curious: what do you think about automation versus humans and how technology is influencing that? NATAN: So, the quick high-level answer that, of course, has a lot of layers into it is what we characterize as the shift from pure automation in the classic sense to augmentation, and augmentation can happen in the individual level. So, imagine, you know, take a couple of examples like someone who grew in manufacturing, maybe they have just vocational education, or just finished high school and, you know, went into manufacturing and grew with the company, and now they gain responsibility. And they understand the process because they learned it on the job, you know, and they're valuable for the company. But, you know, they're still using kind of sticks and stones. As knowledge workers would look into it, and it's, like, how are you doing your job with the spreadsheets, and the clipboards, and whiteboards, and whatnot? And on the individual level, that augmentation is that, if you think about that same person, they probably have a pretty good command of consumer internet. They drive a, you know, a 1,000 bucks smartphone in their pocket. And they can be dangerous. They can do many, many things that you consider pretty high-tech. But they go to work, and it's, like, that's in the locker, which is just nuts, right? So, that's an augmentation that is focused on humans puts the human in the center. But you can extrapolate that. You can sort of think about it at the scope of an organization because if you have many people like that and they have the right platforms, then what happens? You know, that's a sort of a bottom-up yet under-governance type of digital transformation that is augmenting an organization, or, in plain English, let's just change how companies work. This is what Tulip does, you know; we are also a part of that. But I would say that this is what companies are looking for. And it's mostly not for the love of digital transformation. It's mostly, like, to stay competitive because whatever it is, they're trying to ramp up a production line. They're trying to get rid of, you know, issues in their supply chain. They're trying to penetrate new market, whatever it is that is driving that. You know, we're certainly seeing it happening. And that's what I'd say puts the human in the center. I have another point that is related that maybe we can push to you because, like, my answer to your question was, like, shift from classic automation to augmentation. And, you know, in the classic automation sense, you know, if you think about Bright Machines, for example, you know, that build those awesome production cells that package a lot of intelligence and automation and provide flexibility and all the good things that it does, it does still require humans to operate around it. And it's not just, like, turn it on and, like, make me the whatever thing you want me to make, right? How do you see this augment...Like, if we think about augmentation as a trend, how do you see that play into classic automation? LIOR: Yeah, no, it's a great point. And, as you mentioned, Bright Machines it's a company that we built to focus on the back end of manufacturing, basically, where you do the assembly and sub-assembly. And, you know, a couple of the stats that I always love to mention because I think they're crazy: 98% of the electronics in the world is being built by hand. Our fancy iPhone is being built by hand. That will be [inaudible 16:13] about. And our goal as Bright Machine is to move it to use software-defined automation. I love the word, actually, augmentation. I might steal it from you when I'm talking about Bright Machines. Because the reality, even when we operate with Bright Machines, there is things that actually the humans are really good at, and there is things that the robots are really good. Cabling to actually assemble this thing is just not make sense. I will not be able to UPH, the Unit Per Hours, and the first [inaudible 16:43], then the OEE that I need in order to be able to meet the SLA if I'm going to try to put cable. And this is where, inside the line, we will put human stations that will do what human is really good and will work shoulder to shoulders with the robots that will do what robots are really good. NATAN: Yeah, you just stepped into my trap, as they say. LIOR: [laughs] NATAN: [Vocalization] So, bundled in what you just described, there is this mystical term called system integration because one does not have the human next to the machines, and magically, everything just works with the human brain and whatever the machines are doing. One needs integration and the so-called trap. It's not really a trap, as...you know, I think it doesn't work without an ecosystem. You know, that how do you actually do that on time, on money given that, you know, we're making new types of automation and making new types of human augmenting technologies, like, you know, say, like Tulip or others? What makes this new type of ecosystem work? Because if we're doing it in the way the classical automation players and the classical SI delivery, we're just converging to what costs, I don't know, I saw numbers between, you know, $150 and $200 billion spent annually just on system integration in the U.S. These are, like, very old-ish IDC numbers I saw at some point. But it's like, to me, it's, like, insane amount of inefficiency. Who bears that? Well, manufacturers bear that, and then the customers bear, right? LIOR: I think ecosystem is not an easy thing. That's my brutal, honest point of view in that. I think, you know, it's really hard to connect between different companies that's focusing on different aspects of the chain, and, magically, everything works like a charm. And I think this is the reason that system integrators has been around for that long, yeah, is because they're putting the glue between the pieces. I think I'm seeing software will eat, I think, a big portion of that system integration. And I think you will be able to simplify a lot of things with software. I'm obsessive around vertically integrated. I think a lot of the reason that Tesla and SpaceX has been very successful is because they are vertically integrated, and not only that, they are trying to see where they are not vertically integrated, and they are trying to do it [laugh] as well. Naturally, they have deep pockets in order to do it. But it goes back to the Bright Machine things. Yeah, we decided that we will do also that portion on behalf of our customers to just make sure that we are getting our software-defined automation on the floor as fast as we can to meet those SLAs. And that's meant for us to do some system integration along the way to help the customers to give us those very loud deals. But I know it's a fairly controversial point. And I think, historically, you needed a matured ecosystem in order to really transform an industry, and I think it will happen also here in the manufacturing space. NATAN: Yeah. So, I appreciate the brutal, realistic yet somewhat, I'd say, not completely pessimistic point of view. But hey, the internet works, no? LIOR: Yeah. NATAN: So, what's the problem? What's wrong with Open Standards, and APIs, and all that good stuff that makes the internet run? LIOR: Yeah, no. I found that there is a huge difference between bits and bits to atoms. When you start touching the atoms, what necessarily works in the bits world, not meaning, is fully transferable to the atoms world, meaning, yes, that methodology works amazing in the world of internet. But when you start touching physical stuff, conveyors, and materials, and inventories, and factoring lines, and finished goods, and packaging, it's way, way, way more complicated than just integrated bits when you're sitting in your couch. The price will be bigger. So, I'm like; I signed up my life to solve it because I think we are seeing Tesla did it and SpaceX did it. And two companies did it. And you see the price on the other side. NATAN: Totally. But I also see other phenomena. You know, companies like Snowflake suddenly appearing in HANNOVER MESSE. And you think about Snowflake; what are they doing there in a manufacturing premier trade show? You know, it turns out, well, everybody needs their infrastructure and stuff like that. You can argue that. But, you know, mature products like that find their way into the way manufacturing tech is being shaped and built today because that's how we build technology. And I think it lends itself to more internet-driven system integration that we'll see. That's my personal prediction. And, obviously, you know, with my Tulip hat on, we're seeing a lot of this stuff happening already, the way folks are designing, you know, open architectures that they can, you know, mix and match components and things like that. And they're still doing vertical integration to a sort, but, like, mostly, they're trying to take control over their stack and how it evolves, you know, much like we do in any classic IT type of environment. LIOR: I think the reason that Snowflake was there is because of the previous point that OT and IT is getting much closer to each other. When that's happening, you're actually not having data lakes in the world of manufacturing because I'll now start gathering all of my operation data into a central nerve. And actually, this is the Snowflake natural business, and my guess their business is growing fast in that category because of everything we talked about. And the reason, actually, when you're putting Bright Machines line, you are now talking bits, right? Those robots have a control system. So, actually, it's easier to treat it like a software than maybe a traditional way of assembly. I just still feel there is a way to go that you are fighting with physics. You still need something on the lines to actually put the product together. And when you're touching physics on the manufacturing, the narrative of a pure software and kind of the pure software methodologies needs to adjust in the world of manufacturing, not only manufacturing in general, anything -- NATAN: So, we can definitely agree the future will be complex and exciting. LIOR: Yeah, 100%. And the last thing is it is not only complex and exciting is; I do not believe the world will grow from 100 trillion to 150 trillion in GDP without digitizing those industries. And, by the way, I do not believe we can slow down carbon emission without digitizing those industries that are the heavy emissions by technology. And I don't believe that we'll be able to employ 10 billion people without training them to be part of this modern industries again. So, this is not a nice-to-have; this is a must-to-have if we want to stay on this globe for a couple of more hundreds of years. NATAN: Perfect. You almost anticipated my last question, answered it correctly, and rose to the challenge of summarizing what's next in industrial software space as a VC without saying ChatGPT even once. So, congratulations. You're a winner. LIOR: Thank you. Please send me the trophy. NATAN: It's being 3D printed as we speak. LIOR: [inaudible 23:58] something. NATAN: Yeah, by Formlabs, of course. Thanks so much for coming on the show. This was awesome. I truly appreciate it. LIOR: Natan, thanks for having me. It was a pleasure. NATAN: And we'll see you all soon. Thanks a lot.