A Product Market Fit Show | Startup Podcast for Founders
Every founder has 1 goal: find product-market fit. We interview the world's most successful startup founders on the 0 to 1 part of their journeys. We've had the founders of Reddit, Rappi, Cohere, Glean, Huntress, ID.me and many more.
We go deep with entrepreneurs & VCs to provide detailed examples you can steal. Our goal is to understand product-market fit better than anyone on the planet.
Rated one of the world's top Startup Podcasts.
A Product Market Fit Show | Startup Podcast for Founders
His robotics startup raised $400M, his VC fund over $4B—& he ran both at the same time. Here's how he did it.| Lior Susan, Bright Machines & Eclipse Ventures
Lior is the Elon Musk of VC. In just 8 years, his venture fund went from 0 to $4B under management. And while doing that, he founded Bright Machines, which to date has raised over $400M. He's both the CEO of Bright Machines and the Managing Director of Eclipse Ventures.
And he's not building "easy" software startups either. Bright Machines is looking to automate the entire manufacturing process with robots. He launched it with a $179M round and a 100-person team.
Lior is not normal. His story isn't either. You won't want to miss this one.
Keywords
venture capital, startup journey, Bright Machines, manufacturing innovation, fundraising challenges, robotics, automation, customer relationships, product market fit, entrepreneurship, Eclipse Ventures
Timestamps:
(00:00:00) Intro
(00:08:31) Starting Eclipse & Becoming a VC
(00:13:58) How he started Bright Machines
(00:18:43) The First enterprise deal with Flextonics
(00:24:49) The Process of Automation and Assembly
(00:30:25) Making a Machine as Reliable as a Human
(00:34:44) Bright Machine's Struggles
(00:36:56) The Business Model of Robotics
(00:39:49) Finding Product Market Fit
(00:40:37) One Piece of Advice
Lior Susan (0:0)
I will tell you what, if you tell me in 2015 that in 2024 I will manage $4.5 billion and I will be a venture capital, I will laugh in your face.
Pablo Srugo (0:08)
So would most people.
Lior Susan (0:10)
I had the privilege to meet both Elon and Jensen. And Elon told me a couple of times, they had four or five times a Tesla with weeks in cash. And he was like begging the state of California to give him a loan. And without that loan, There would be no Tesla today. I said there are other things that I think people underestimate. We got super lucky. Luck is a big portion of this business. We got super lucky that our thesis became mainstream and the largest companies in the world right now are full-stack companies, right? Microsoft and Nvidia, Tesla and Apple, they're all companies that, I mean, if you're seeing where Meta is spending most of their money is on hardware, not actually on software. So most software companies went to hardware. So we just got lucky that the thesis became mainstream. I had the privilege during this job to meet the CEOs of the largest companies in the world. And they do not need to give me, I'm nobody, they do not need to give me their time. And when I heard from them, their challenges around manufacturing and saw in their eyes what we are describing that they are doing, how it's going to impact the largest companies on the globe, I knew that we are touching a product market fit.
Pablo Srugo (1:26)
Welcome to the product market fit show brought to you by Mistral, a seed stage firm based in Canada. I'm Pablo. I'm a founder turned VC. My goal is to help early stage founders like you find product market fit.
Well, Lior, welcome to the show.
Lior Susan (1:43)
Thank you, Pablo. Great to be here.
Pablo Srugo (1:44)
Dude, so the first thing I want to ask you right off the bat is, you know, it normally takes like 110% of your attention to do a startup to get any sort of traction in a startup. And then here I'm like reading about you and I'm like, okay, this bright machines thing. Like it seems to be doing well. You raised like over $400 million. Your CEO, co-founder there and then got Eclipse Ventures, which is like 4 billion AUM. We just raised like $1.2 billion last year and you're the CEO and founding partner there. What's going on, man?
Like just like, how are you making this happen?
Lior Susan (2:15)
Hey, You know, Elon makes everyone look so easy. I mean, you know, SpaceX, Tesla, X, who I am that I'm going- That's what I'm telling my wife. It's only two day jobs. There is other people that are having seven, eight day jobs and they're frankly doing well.
Pablo Srugo (2:29)
That's crazy, man. Maybe let's start a little bit on Eclipse, I mean we'll talk mainly about Bright Machines today, but I think one thing leads to the other. Even Eclipse seems to be kind of this big zero to one moment. If you look at a lot of even the big funds, I mean not the big funds, but the early stage stuff, they started a $10 million fund. For us at Mistral, the first fund was 20 million. Your first fund was like 125 million, at least according to Crunchbase. What's the story there? How did you get that off the ground?
Lior Susan (2:53)
Yeah, you know what, in some way, the story of Bright Machines, myself, Eclipse is very much connected to each other. I was building companies inside Flex and the original Bright Machine started there. We'll talk about it later. And when I left Flex because I want to continue to build companies, just wanted to do it on a standalone platform rather than inside Fortune 500, I knew that I need capital in order to do it. And when I always say the power of ignorance, I did not know what limited partners were. I did not know how hard it was to raise $125 million to your first fund. I did not know all of that. I'm like, great, I'm an entrepreneur. I raised money for my companies before. I'm sure I'll be fine. It was much harder than I thought.
Pablo Srugo (3:40)
Walk me through maybe then your story. You mentioned Flex, Flextronics, which maybe you can tell us a little bit more about that company. Obviously the name is well known, but what about before that? What was your kind of story before that?
Lior Susan (3:52)
Yeah, I grew up in Israel in kibbutz community, joined the special forces for my military, was there a bunch of years. Left in 2008 to start a company with my young brother. We did a software defined network for the mobile carriers. Cisco bought the company in 2012 for 475 million and moved me to the Bay Area. Now I was not the brain, it was my brother. So I was just here for the integration.
And in one of those meetings, I met the CEO of FlexTronics that was the main contract manufacturing of Cisco. And Mike convinced me to join him and built for him the digital transformation team that will incubate tech companies inside Flex instead of going back home. And I joined him for one year. I was there for more than three. And yeah, if the people that don't know who is FlexTronics, it's like the American version of Foxconn. So about 300,000 people around the world doing electronics manufacturing.
Pablo Srugo (4:52)
What made you decide to join- like most people who sell a business either just go to the beach or start another business or maybe sometimes start a fund, but few will join another company and work.
Lior Susan (5:00)
Yeah, actually, my plan was a beach. Now, not like many Israelis that after the military that go on a long trip, was after a long military, I immediately jumped into Intuosel. And when we sold the company, I just got married to my wife and I told her, hey, we're going to go, I want to kitesurf in Brazil for a year. So this will be just a few months in Silicon Valley on our way to Brazil.
Pablo Srugo (5:25)
She's still waiting eh?
Lior Susan (5:27)
13 years later, she's still waiting.
Pablo Srugo (5:31)
That's right, dude. What did you see there that made you want to change your plans?
Lior Susan (5:34)
I'm a full stack person by heart. When I say full stack, I'm not talking about actually software full stack. I'm talking about hardware, data, AI, computer vision, software, et cetera. And when I met Mike, he's like, Hey, we have one of the largest manufacturing on earth. And if you are going to come here, I will let you use this massive supply chain, manufacturing and logistics beast that we have that we have over here to build companies. You know, my head spun super fast. I told him, yes, very, very fast. So I did not think too much, it is just a gigantic playground. And as a guy that liked to build,it was a dream.
Pablo Srugo (6:13)
And what was it that you were supposed to do there? What did you end up doing?
Lior Susan (6:16)
Yeah, I created this team internally that is called Lab 9, still running there. And basically our team was in charge to identify which opportunities there is inside the big flex in order to go and build small special forces tech team to go and solve this problem. Manufacturing automation, supply chain software, predictive maintenance, new MES, manufacturing execution software, new shop flow, the ability to interact with our customers, connectivity. So we basically worked on thesis and we went out and hired teams and built those companies internally to go after those thesis.
Pablo Srugo (6:55)
And so you do that for three years and is this Bright Machines, is that connected to the time you spent over there?
Lior Susan (7:03)
Yeah, you know, Mike, I flew to Zhuhai, one of the largest facilities of Flex in China. There's about 80,000 people on that campus, to give you a sense. And here I am stepping into the floor and thinking I'm going to see all of these robots building Apple product and Cisco product and Johnson and Johnson product. And I stepped to the floor and I see long lines of people standing next to each other. And I'm asking the plant manager, where's the robots? He's like, there's no robots. And I'm like, you want to tell me that we are building the most advanced electronics on earth by hand? And he's like, yeah, robots are not flexible enough for what we do.
Pablo Srugo (7:45)
This was when by the way, like a decade ago?
Lior Susan (7:47)
2013. And here I am telling that guy over there, I'm going to build a company that's solving this problem. And he's like, Mr. Lior good luck. So I was like, okay, let's go. build the thesis, hire the core team and start working on that while I was at Flex.
Pablo Srugo (8:04)
That was like within Flex, like this special forces kind of project?
Lior Susan (8:07)
Exactly. That way we call it originally AEG, the code name, Advanced Engineering Group. We worked on that thing internally. I left in the beginning of 2015 to start Eclipse and then in 2017 came back to Mike and the Flex boards and said, hey, I want to carve out that team to a new company that I will create it called Bright Machines. And that was the birth of Bright Machines.
Pablo Srugo (8:32)
So walk me through even just starting Eclipse, like what makes you decide you're working in Flex, you've got like all like that full stack that you're talking about, and you've got this specific project, which ended up being obviously leading to something massive. but then you decide to go start a fund, why?
Lior Susan (8:47)
Yeah, it's a combination of three things. A-Tesla was starting to take off. I saw the vision of what I believe the world is going to be, and it's basically a technology company building a car. It's actually not a car company that's using tech. And this is the type of companies I love building. And I also saw that the public market value this company as a tech company, as a software company, although they have- back then negative gross margin until today mid teens. And it's because the size of the market that they are going after is incredible. It's a trillion dollar in US and the team that they build over there was able to build a full stack company in order to go and transform mobility. And I just, my vision is we're going to see a lot of Teslas and SpaceX and Nvidia is being built across those physical industries like energy, supply chain, defense, et cetera.
Pablo Srugo (9:41)
Like these are companies that are like tech first in terms of their mindset but go after and deliver like from a business model, they don't look tech first, but in terms of mindset, they are.
Lior Susan (9:49)
Exactly. You know, Apple is not a subscription, didn't start as a SaaS company and Nvidia did not start as a SaaS company. And for short, Tesla, it's not a SaaS company. So not what Silicon Valley really like in the last 20 years, but this is where's my passion. And I saw that Tesla started taking off and I saw that I think the market is shifting between geopolitics issues, shortage of labor, product changes, the need of building those companies growing significantly. And as much as I love Flex and Mike and the team, I didn't feel like the right thing for me is to build those companies inside of Fortune 500. I thought the right structure should be a fund to build those companies.
Pablo Srugo (10:30)
And so you wanted to do these kind of hardware meets software, old tech, that's where you're looking to fund. And what was the...Like that fund 125 million, was it like seed stage, early stage of what kind of checks were you thinking you were going to write? When did you think you were going to play?
Lior Susan (10:49)
A super early stage, either us building the company by ourselves or writing or leading seed series, say that was the thesis. I remember that I was like, we will write $5 million into each company and it will be enough for those companies. I was wrong and about all of the magnitude, but it's fine. And yeah, we kind of left in the beginning of 2015 and in April started the fit.
Pablo Srugo (11:11)
And what led to such an incredible growth in AUM? just, you know, this is more about founders, but like this stuff's obviously adjacent and interesting. Like, but just high level, you know this, others might not. raising a fund is really hard. I think the odds of raising a second fund are like 50%. Like, you know, the number of firms that raise five funds, I don't know what it is, but it's like 10%. Like these are really small numbers and yeah, if you're doing well, the funds sizes go up. But to start with a hundred million dollar fund, you know, less than a decade, like eight years later ,have 4 billion under management is not normal. So like, I guess I'm curious, like what happened in your case that allowed you to do that?
Lior Susan (11:49)
I will tell you what, if you tell me in 2015 that in 2024, I will manage 4.5 billion dollars and I will be a venture capital, I will laugh in your face.
Pablo Srugo (11:59)
So would most people. So would most people. Yeah.
Lior Susan (12:02)
I'm like, I never thought about myself as a venture capital till this day. And I didn't really imagine me raising fund over fund and building a franchise that was not the goal at all. My goal was like, I want to build a platform that can build a lot of companies in the digital transformation. Well, the Chairman of Flex back then told me the best thing you can do if you want to do it is to start a fund. And I'm like, starting a fund, shit, I never invested a dollar in my life. But you know, I'm like, I'm an entrepreneur. It's probably like starting a company, right? I will find some investors and I will figure out the structure, how hard it can be. They're not really hard, but this is hindsight 20/20.
Pablo Srugo (12:45)
Did you have some really big winners in that early fund that kind of created a tailwind for later?
Lior Susan (12:51)
Yeah. I've said there are other things that I think people underestimate. We got super lucky. Luck is a big portion of this business. We got super lucky that our thesis became mainstream and you know, the largest companies in the world right now are full-stack companies, right? Microsoft and Nvidia, Tesla and Apple, they're all companies that, I mean, if you're seeing, well, Meta is spending most of their money on hardware, not actually on software. So most software companies went to hardware. So we just got lucky that the thesis became mainstream. I also got lucky that I was able to attract much better talent than me to Eclipse. And I do think you need to be lucky with the people that you choose. And then we got lucky that we built a couple of companies that turned out being good outcome. We sold the Six River 450 million to Shopify super fast. We sold Kindric to Ocado. We took Owlet Public. taking another company, Public now, out of that fund. We have Augury doing high teens close to $100 million in ARR in that fund. So we were just lucky.
Pablo Srugo (13:57)
So I understand these are like, this is traditional VC or did you do some kind of incubation where you would come up with some ideas and find the teams around them.
Lior Susan (14:03)
Actually, the first company that we built was Bright Machines in 2017. That was a purely incubation. From 2017 until today, we build about three companies every year. It depends on the year. say two to four.
Pablo Srugo (14:18)
So then walk me through Bright Machines. Because even that, like I understand, the idea was seeded before, right, in your flex days. But for a venture fund that now you're two years old, you've probably already done quite a few investments, then decide to get into the incubation game. I mean, it's quite a leap, like as much as, you know, like I work with, like all the VCs, they all have ideas, right? But then you're actually going to say, okay, I'm going change my model. I'm going to get in it. Like, I'm going to do the work. So like, how does that thinking, you know, kind of happen?
Lior Susan (14:48)
In a very traditional me way, I'm not thinking too much. I'm just going and doing it.
Pablo Srugo (14:53)
It seems to work well for you.
Lior Susan (14:54)
I don't know, you know, not always. had, you know, not always. ask my wife about it. We were obsessed, a lot of us here came from the manufacturing background. Greg Reichel led the Tesla operation and manufacturing team, myself, Charlie, and G10 that came from Rivian and Tesla and SpaceX and others. So we always being passionate about manufacturing and we are very thesis driven. love as all of us as an operators, we write investment thesis, many tens of them every year. We're just getting obsessed by idea and we will go and write for ourself investment thesis. Even if we never find the company in that space, we're just obsessed and I put a memo about electronics manufacturing and I told them in my lifetime, people are going to design iPhones and GPUs in a different way and they're going to manufacture them in different way in the way that has been done in the last 50 years. And we kind of talked about the thesis and I say, this is not fair just to be upfront. actually created similar theme on that in my previous life at Flex. I went to see McNamara, the CEO and I say, Hey, I want to carve that team that they originally built into a new company. He said, no, you lost your mind. we're never going to give you
Pablo Srugo (16:08)
*Laughs* What's the upside for him? Yeah.
Lior Susan (16:11)
Yeah, he's like, no. And I convinced him why it's a great idea for Flex and for us and for those people and everyone.
Pablo Srugo (16:18)
Was the idea that they would get, I guess, a piece of that new co or something like that?
Lior Susan (16:22)
Exactly. And that other point for me to him is like the company will require many hundreds of millions of dollars and you at Flex cannot spend that money on AI engineers and robotics engineer and computer vision engineers. This is not the margin profile that Flex has.
Pablo Srugo (16:36)
And why did you decide– oftentimes when this happens, the typical playbook at least is you might create that team. kind of have the idea, you create the team and then you quickly put a CEO on it, but you kind of stayed at the head of it. Why did you choose to do that?
Lior Susan (16:50)
Yeah, so it's not always the case when we build. We’re also staying the CEO for a long time. I think the goal here is to find much more capable people than us to take it. And we're doing that very often. I will say that I am incredibly passionate about this problem. It was hard for me to give away, I'm working on it, to give away the ROM to someone else.
Pablo Srugo (17:13)
So you finally convinced them you have this team. How many people?
Lior Susan (17:17)
About 170 people they want.
Pablo Srugo (17:19)
Oh wow. Okay. I was thinking like five or 10 people. Okay. So it has to come with like that, because the first round was in the hundreds of millions, right?
Lior Susan (17:30)
180 or so.
Pablo Srugo (17:31)
Okay. So it had to be that way because 170 people, mean, you need a lot of money.
Lior Susan (17:34)
You need a lot of money. And the other thing, it's a lot of chicken and eggs. If you are going to start with two guys in a presentation, none of the largest companies in the world is going to let you touch their mass production product. So actually if you want to create the transformation, You almost need to have a certain scale out of the gate.
Pablo Srugo (17:52)
How do you get that chicken and egg going? First step, Flextronics. Yeah, okay, fine. We can do this new code thing. You can take these people. Okay, you got the people. You need the money and you need like the, I guess, design partner, just something like that. How do you get those other two pieces?
Lior Susan (18:04)
So what I did when we carved out, I also agreed with Flext that they will be our first design partner as a customer. So we signed a multi-year agreement with them as the first customers. With that deal that we signed with them, I turned to an investors and I say, Hey, I'm going to create the next Foxconn or whatnot. It's going to be a hundred billion dollar company. Hopefully one day. And for that, need a lot of money. And by the way, I have this amazing team and amazing customer and I got the investors and you know, kind of start running fast. And I got a lot of calls from my LPs of like, what the heck are you doing? You're not supposed to build companies supposed to invest. And I told them, listen, I, this, this is who I am. I'm just incredibly passionate about building companies. I cannot stop myself.
Pablo Srugo (18:50)
When you talk about that deal with flex Tronics, what is that deal? Right? On the one hand, in terms of revenue, what does it look like? But also what are the kind of milestones or checkpoints that are part of that kind of construct? How do you set that up?
Lior Susan (19:04)
I would say I never did a carve out before. So it was like a learn. I was flying the plane while learning it. but you know, I figure out that the way that you do it is you have the people. So there is a lot of work with the HR figuring out who these people are, where they best, entities, all of that shit. There is all of the IP, everything that these people developed till today, source code, IP, patents, trade secrets. There is all of the equipment. We had tons of lines and robots. So like, where is the equipment? How much do you value the equipment? And then it was a lot of negotiation with Flex of what's going to be the valuation of the new company because they are going to get 10 % or whatever, can't remember, of that new company. So I spent a lot of time with lawyers and negotiating these things, but it was my first carve out.
Pablo Srugo (19:50)
Because it's almost like an acquisition. I mean, you're funding it with new equity, but you're almost like acquiring these people in this team and these equipment from Flextronics.
Lior Susan (19:58)
Exactly.
Pablo Srugo (19:59)
And then from a design partner perspective, like what did that business deal look like?
Lior Susan (20:03)
We signed, we actually got, I think we did a really good job understanding that we need Flex to provide us a couple of low tens of millions of dollars in revenue every year for the next couple of years. So we can just focus on them rather than go to them trying to sell to the whole world to help us get the product roadmap of where it needs to be. Cause it was clear to me that those guys and ladies worked on that problem inside Flex, but the vision that we had where we wanted a sprite machine was not exactly the same. So we actually need to take that core. And it was obvious that we will need couple of good years of actually building what we call smart skills. And basically that's the AI brain of those robots, all of the control, the computer vision, the cloud, the data, the recipe as a code, our digital twin, it was a lot of developments and I did not want to spend time on a go-to market in the first couple of years. So Flex was just a great partner for us as we were building that in.
Pablo Srugo (21:08)
And what,- they're paying tens of millions of dollars in revenue. You know, you have hundreds of people, was there already a product kind of that was in market at production level and you just wanted to build on that or were they-
Lior Susan (21:19)
No, it was really early. Basically what we carved for them was what we call BRCs or bright robotic cell. And it's a very flexible robotic cells. That's essentially what those people build internally at Flex and they will just replace manual processes with those BRCs. And we took the hardware and on top of it built a ton of software that allowed us to now not only replace people with single skills, but also to transform how the manufacturing floor looks like from an end-to-end process. And also to give those robots the same granularity of flexibilities as humans have, because the lines are changing what they are building all the time. And this is the opposite of automation, historically.
Historically, robots was building just one product again and again and again. And we here in the electronics manufacturing doing once one floor of manufacturing can build many tens of different products every year.
Pablo Srugo (22:18)
And was there- they're going to pay tens of millions of dollars in for X amount of years. are they saying “then this year I need this many robots or I need this percentage of automation done” or what was like the thing that they were paying against?
Lior Susan (22:29)
Yeah, we did. It's a great question. We actually did not know. So I was the hard part and we almost agreed that we will partner in X amount of dollars and then during the year we will identify through an SOWs what will be the specific needs of each of their factories. They have 200 factories around the world. I would say it did not work perfect. It's actually was really hard during the first couple of years because some of those sites had needs that we did not want to work on because it was not aligned with our roadmap. But then, know, flex say, but hold on, but we commit, if you want the dollar, you will need to do it. I will say overall, it was a positive partnership, but it was not perfect.
Pablo Srugo (23:12)
But it was almost based like 99% then on trust, right? Like there's no way somebody from the outside could come in and pitch this to this CEO and be like, yeah, okay, let's do it.
Lior Susan (23:21)
And it works amazing in the first two years and then Mike left. It's got much harder with the new leadership, nothing about those people, but it's more like… I did not have the same trust I had with Mike, but yes, you're a thousand percent right. It was pure trust game.
Pablo Srugo (23:37)
Because there's no way you're getting— they all got upside ultimately, but the amount of risk they took upfront was pretty massive, right? And like the amount of leeway that they gave you in the way that you described that, that business deal, that contract, you know, the only reason that makes sense, I think like looking at it from the outside is he just thought, you know, you would make it happen. You and that team would ultimately, you know, put something together that was valuable.
Lior Susan (24:00)
And remember, it's a $30 billion revenue company. They don't care what the equity is going to be worse. It's anyhow below the line for them. He was doing that because mainly, I believe, you need to ask him, because he believed this industry is going to change. And he felt that the Flex will be part of the company that will change these industries. It's worth positioning them in a way that they will not feel like they were GM or Ford and Tesla just being created and they were not part of it.
Pablo Srugo (24:30)
And by the way, maybe you can walk me through just an example of, I'm totally an outsider to how these chips are being made. You talk about these manual processes and in my head, I have an image of people putting things on chips or whatever. Walk me through maybe an example of the sort of processes that you're automating these days.
Lior Susan (24:49)
Yeah, maybe let's just take this AirPods case as an example, not because we're building it, it's just that it will be an example. Here in this AirPods case, you have a couple of parts. You have the plastic housing, that will be an injection mold. You have multiple of PCBA, that will be the electronics cards that's sitting in this thing. And maybe you have some metal or magnet here and some lighting. The way that it works today, Apple will go buy chip from TSMC. They will take those chips, will ship them to someone like Foxconn. Foxconn have two sides of the line. The first one will be what's called SMT. And this is those high-speed machines that will take those chips and we'll put them on the board. The front end of electronics manufacturing is pretty much fully automated already.
Pablo Srugo (25:44)
So this is putting chips on the PCBs.
Lior Susan (25:46)
On the PCBs, exactly. Now you take out PCBs and now you need to take the plastics piece, the PCBs, the magnets, the lighting, It's called the backend, the assembly and the sub assembly. That's 98 % human based today. And my vision was like, want the front, the backend to looks like the front end. I want to create those robots that will.
Pablo Srugo (26:07)
And the reason for it, I would assume is that putting chips on circuit boards on PCBs is a bit more, is a lot more repetitive than, than what you call the back, the backend manufacturer. What's the reason that that was automated so much faster?
Lior Susan (26:19)
Good question is the variety of what's happening in the front end is much less. And then to the day, you are always going to pick a chip and put them on a board. That's it. That's the job. And you need to do it really fast. Then things in the backend.I mean, think about this Mac book that we are talking from these airports and a GPU. They all look so different than each other, meaning the assembly of those products looks very different.
Pablo Srugo (26:47)
I see. And so what, what does that world look like when you start the company, you know, eight years ago or so?
Lior Susan (26:53)
For all of the humans, basically it's like the majority of this industry is close to a hundred percent will be people. And it will be a labor arbitrage game, meaning you will try to find the places that have a cheap labor, put a lot of them next to each other, build those products, put them on a boat, ship them to your customers and do it again and again and again.
Pablo Srugo (27:16)
And that's still split. Like if you take the air pod case as a great example, everybody's still just doing one piece of the puzzle and moving it to the next person or how, how is that kind of usually split up?
Lior Susan (27:26)
Usually in the case of the airpods , because you know, it's a very large companies like Foxconn and others, they do a lot of those processes end to end, but it's they will basically will have one factories doing the injection molding one factories doing the SMP one factoring is doing the assembly. They will build the finished product, put it on a box and we'll ship it on a boat to whoever's the customer.
Pablo Srugo (27:49)
But I mean, within the assembly, I imagine like all these humans and they're all just doing one part of the assembly. They're like clipping this, moving to the next person in line?
Lior Susan (27:57)
Yeah, yeah, sorry, I did not understand the question. Yes, basically you will have these extremely long conveyor belts and you will see people standing next to each other and across the things and each of them is going to do one task. Someone just putting that screw, someone just putting that label, someone flipping the product, someone, yeah, like scanning. It's just one by one they will do those processes.
Pablo Srugo (28:21)
Now with Bright Machines, the goal wasn't so much to take, you know, one piece, let's say flipping it and create a robot for that. It was to create robots that could do, that could be reprogrammed, I would assume, and kind of for this AirPods, you're to do that. But then for the next thing, you're going to do something else. That was kind of the hypothesis or the goal, I should say.
Lior Susan (28:37)
Exactly. We wanted to do the process end to end. And in our line, there is still some people we are working in a hybrid because there is things like cabling that maybe will not make sense to do with robots. And he wants the humans to do it. But the other issues about humans is we don't collect data when humans build products. When you start using robots, you collect a lot of data that maybe you never had around how much force you're putting and how that's potentially hurt the PCB and it will fail down the road. So what we start finding when we will start creating this robotic solution for multiple of steps in the process, we start getting insight that those insights will be relevant to the customers to how they should design the next product.
Pablo Srugo (29:24)
So you're not just - The ROI isn't just cheaper costs to produce. It's also around product quality and things kind of further down downstream.
Lior Susan (29:34)
We take the whole data from our manufacturing lines and we tell our customers the supplier quality of the parts they're buying. We tell them the quality of the product, how many we produce, what was the cycle time. We feed the whole data back to our design for automation engine. So while they're designing a new product, We have a generative AI engine that's learning constantly to hopefully one day we'll be able to build the most perfect product. So you're basically catching both of the barbells, from the design of the products to the manufacturing on the products and just connect the data across the board. And it's incredible. We are finding insights that is just incredible.
Pablo Srugo (30:17)
So going back to those early days, now that I understand kind of the scope a little more clearly, You're working with flex Tronics, like, and I know the concept of MVP doesn't make as much sense here, but you still have to somewhat constrain the promise. They like, this is the part that we're going to solve first. How do you do that? What do you go after first?
Lior Susan (30:36)
We start very simple of forget about end to end process, forget about the foundational model for robotics. We start with, can you be as reliable as a human can be doing electronics manufacturing tasks? Just start there, just putting screws, just putting themes, just putting heat sinks, just putting CPUs, just do welding here, just do gluing, scanning, the tasks that you will do very often in manufacturing. And the hardest thing about Bright Machine is the people that our comparison is not in other startups, so it's a big company, it's humans. And actually humans are pretty good. And then the return of investment needs to be very attractive for these people because they know how much a human costs them. Now they're struggling to find those people. There is incredibly high attrition. They want to onshore manufacturing so they cannot have the same amount of people they had in China. But the gold standard is this amazing machine. It's called human.
Pablo Srugo (31:40)
And so what you're doing is you take a subset of activities like processes, like let's say putting screws, flipping something, and you go after those and you try to match or exceed, let's say human performance.
Lior Susan (31:52)
Exactly. So we started in the first couple of years just matching. Can we do the same unit per hours as a human? Can we do it in the same quality as a human? And then we start, we actually can do more per hour than a human. we can actually do in a higher quality in humans. And then we say, OK, can we do now two processes? So the one robot will put screws. The other robots will do pick and place. Now we measure the cycle times instead of two humans, two robots, and would basically grow that floor as we are progressing with the roadmap.
Pablo Srugo (32:26)
You know, one question that comes to mind is like, so I had a hardware startup before, it didn't end up working, but I'm familiar with some of the issues of building hardware, working with hardware, just like how frustrating that could be sometimes when things aren't working. In your case, like the opportunity is massive, but also the stakes are really high because … You started this deal, you didn't start from zero. You started with a lot of people, all this money, all this stuff. Things really got to work. Were there times in the early days where, you know, whether things were very behind schedule or whether there were parts that you're like, shit, I wonder if we're actually ever going to be able to do this. You know what I mean? Like, is this ever going to be realized or was it just kind of like pretty, pretty smooth sailing?
Lior Susan (33:06)
I still have that feeling. So, you know, it's like we rarely meet the timeline. And we almost always late to what we commit or we'll do. This is incredibly hard. But the thing I'm getting excited about is just our fleets of robots in the field is growing through hundreds of hundreds of hundreds. And we start showing things that the customers never imagined we will be able to do. Like we find problems that they never thought that they have, or we help them optimize their process in a way that they generate so much more servers for AI or whatnot our customers are building. So I think we are now in the inflection point and we now the platform is robust enough. But yeah, I had many of those thoughts along the way. I also had some thoughts,you know, hey, this was a good idea to start with this type of a scale rather than me starting the companies like I usually do with five guys and a dog. And until this day, I'm not sure that I'm exactly certain about the answer.
Pablo Srugo (34:16)
Were there any like, again, because the stakes are so high, the burn is high, and you kind of you had COVID that happened through this. I don't know what that effect was.
Lior Susan (34:25)
Oof. COVID was almost killed us. COVID was brutal. We're a full-stack company. We work on the robots with our software and our ML team. It's… You can't do those things from remote. Now on the flip side, it shows vulnerability of humans in manufacturing. So the need of robotics went to the roof post COVID, but COVID was really hard for us.
Pablo Srugo (34:51)
Were there times where you had, you were really tight on cash, really tight on runway?
Lior Susan (34:55)
I don't know how much we want to share in this broadcast, but we had the situation of weeks.
Pablo Srugo (35:05)
Wow.
Lior Susan (35:06)
I'll tell you one more thing. At the peak, we were 700 people and now we are about 200. So, you know, it's not, and we figure out that we actually need less people, but we need people to be more in our HQ in San Francisco than maybe we had around the world. So it's not always rosy and up to the right.
Pablo Srugo (35:26)
When were you 700? In what year?
Lior Susan (35:29)
We were 700. We almost spacked the company about two and a half years ago. And kind of in the last moment, I decided that we're far away from being public company ready. And I pulled the plug and said no to a very large financing and needed to do some significant rifts. That was not fun.
Pablo Srugo (35:31)
Probably a good idea to not go down this path with hindsight in mind.
Lior Susan (35:55)
Hindsight looks brilliant.
Pablo Srugo (35:57)
That's right. But it's funny, the thing about weeks resonates just because that's one of my biggest… let's say learnings or whatever you want to call it, which is like just how many companies that are either successful or like from the outside look like very much on the path success went through times where, you know, he could have flipped the coin and it'd be dead, right? And there's another world where this company doesn't exist anymore.
Lior Susan (36:20)
I had the privilege to meet both Elon and Jensen and Elon told me a couple of, they had four or five times at Tesla with weeks in cash and he was like begging to the states of California to give him a loan and without that loan, there was no Tesla today. And Jensen likes to reminds me that Nvidia was an overnight success of 32 years of hard working. told him, you know, a very first call, told him, God, I'm only seven years into it. Do you think I can do it faster? And I, good luck. If you can, let me know how.
Pablo Srugo (36:56)
That's wild. And you know, another question I have is around just the business model, right? So you mentioned you have hundreds and hundreds of robots. Is this like a CapEx thing? They pay a few million dollars for robot or is this like a kind of robot as a service type of model?
Lior Susan (37:09)
We do both. We have a CapEx model plus subscription on the software as well. have robotics as a service model. And tell you the truth, I literally just talked with the board last week on that. I have yet to figure out the right business model, although we are doing many tens of millions in revenue. So, you know, I think still we are learning what should be the long-term business model.
Pablo Srugo (37:32)
And is that more from what makes most sense for you or just like in terms of what resonates with customers? Like do you find some customers prefer the robot as a service model, but others are more used to just buying equipment or how's the market respond to that?
Lior Susan (37:44)
I think the market is careless because our customers are fortune 100. They have so much cash on their balance sheet. So I actually less, I think my point there is less about the market is just, I believe if we are successful, someone is going to his CAD to design a new AirPods. While he is designing that AirPods, our software will run on the background of the CAD and tell him what he needs to change in order for that things to be designed for automation. Once he's done, his manufacturing engineers is going to get a file that will, in our simulation tool that's it on Omniverse, you will be able to design the lines. Once he's done, you will be able to press a button and a new file, we call it recipe, is going to go into 30 factories around the world that build your product with our robotics. And the robots will reconfigure themselves and will know now how to build this new product. And once they are building this product, they are going to ship you all of the data back of how they are doing. We are far away from what I just described, but we are making significant progress. And I think if we will do what they just described, my guess our business model will look different than what it is.
Pablo Srugo (38:57)
It has to, it has to, because you think like the way I think about your business today, and I am probably wrong, but just from the outside, you're like, okay, you got robots, like that's the core thing, right? But when you, when you talk about the way you described it, the value offering is very different than a robot. And so you got to price it completely different.
Lior Susan (39:14)
Exactly. The robots for us, it's people think about as a robotics company. We are not a robotics company. We are next generation manufacturing company. Robots are part of our stack. Exactly right. So yes, so TBD on the business model.
Pablo Srugo (39:30)
Cool. Well, listen, we'll stop it there. I mean, you've raised at this point, like I said, over 400 million degrees. Last round was $100 million, just over $100 million. A few months ago, you had the SPAC thing that almost happened. So I think there's a lot of crazy things that haven't happened yet, but I expect to see this name Bright Machines for sure in the future. Let me just ask kind of the last two questions that we always end on. The first one, And it's weird one for this one, but like, when did you feel like you had true product market fit?
Lior Susan (40:02)
I had the privilege during this job to meet the CEOs of the largest companies in the world. And they do not need to give me, I'm nobody. They do not need to give me their time. And when I heard from them, their challenges around manufacturing and sawing their eyes, what we are describing that they are doing how it's going to impact the largest companies on the globe. We knew that we are touching a product market.
Pablo Srugo (40:32)
And then the last question is if you could go and especially because you've worked with so many other founders, given what you do at Eclipse, you know, what are some of the most common pieces of advice that you find yourself giving to founders in those really early kind of zero to one stages?
Lior Susan (40:50)
You know, it's the cliche about being customer obsessed and people are telling me that all the time. And then when I'm digging in, they're not. And I think, know, Flex was an incredible design partner for us. And then one of the largest hyperscaler. Without these people, I think we will not be able to understand what we need to build well. This industry, especially physical industries, if you just build it does not mean that they're going to come. You need to constantly iterate on the product and we are still constantly iterating on the product and you cannot do it without having a few very strategic customers that are Fortune 100, 500, and that you are all truly listening to what they have to say about their business.
Pablo Srugo (41:39)
To you, what is it? Customer obsession gets thrown all the time and it's one of those things where it's kind like what you just said, there is no founder that's going to be like, I'm not customer obsessed, fuck that. You know, like they all do , so what does that look like to you? What does it really mean?
Lior Susan (41:52)
Till this day, some of my top engineers, our customers will come and say, I want this or that. And they're like, they don't understand. I'm like, stop it right now. These people are fortune 10. They will understand their business. We are nobody. So even if we disagree, we are going to take their feedback and we might decide to go elsewhere, but we are going to take that feedback and we will never say, “They do not understand what they're talking about”. So I think it's just embedded in the culture, especially when you're in a deep tech engineering organization that 80 % of my team are engineers and PhDs and smart people around AI and computer vision, software and robotics. It's critical.
Pablo Srugo (42:38)
Perfect. Well, Lior, thanks so much for jumping on the show. It's been awesome.
Lior Susan (42:43)
Pablo, great to be here. I just gave you content that you liked so much, you actually listened to the end and guess what?
You didn't pay a single dollar. Not only that, I didn't even put any ads in your face. So you just got a bunch of content for free. And now that I've delivered that value, I'm asking for something in return. Open your app, open Apple podcasts, open Spotify, open whatever app you use to listen to this and hit that follow button. It's actually going to help you because it's going to help you make sure you don't miss out on the next episode, which you liked so much that you listened to the whole thing.
