Episode 432 of The VentureFizz Podcast features Johannes Galatsanos, CEO & Co-Founder of Diffraqtion.
I’m incredibly lucky to host this show because the guests I interview are truly out to rewrite the rules of what’s possible. Johannes is a perfect example. He brings over 15 years of deep expertise across AI, quantum tech, and operations across a career that includes lots of deep research in academia to executing massive, corporate AI initiatives at global giants like Novartis.
So, what does someone with that background do to push the absolute limits of technology?
How about building a space company that is creating the world’s first quantum camera to help satellites and machines see further and think faster?
By blending quantum photonics with cutting-edge AI edge-computing, Diffraqtion’s technology enables satellites and telescopes to bypass traditional lens constraints entirely—delivering 20 times higher resolution and 1,000 times faster processing speed, all at a fraction of the cost.
As you’ll hear from this interview, there are a countless number of use cases and industries that are perfect for Diffraqtion’s technology, which puts them in an ideal spot to build a massive anchor tech company in the Boston startup scene.
In this episode of our podcast, we cover:
- Johannes’ background growing up in Greece and how his early fascinations with computers and astronomy led him to the world of deep tech.
- His professional journey across academia and roles within industry.
- The mechanics of how a quantum camera actually works and why their underlying tech acts as a massive “moat of defensibility.”
- The origin story of Diffraqtion and how he met his co-founders, plus the latest with the company.
- The importance of building in public.
- Johannes’ unconventional approach to fundraising, including using a scrappy, rolling funding model to drastically reduce equity dilution.
- Why he firmly believes Boston is the right ecosystem to build a global space intelligence company.
- Plus, so much more!
Transcript
Keith Cline (02:37)
Johannes thanks so much for joining us.
Johannes Galatsanos (02:40)
Thanks, Keith. I’m glad to be here.
Keith Cline (02:42)
I’m excited to talk to you because we’ve got so much to cover. What you’re building is extraordinary. So it’s like one of these things where like, how do you even get started? So we’re going to talk about that. It involves space. So that’s super cool too. But when I was looking through your background, obviously AI is on the tip of everybody’s tongue as it should be. Yet when I look at someone with your profile,
AI is something you’ve been doing for a very long time and we’re to talk about your background, your history, and we’re going to talk about what you’ve achieved in that. But you also are a unique person that has this understanding of quantum computing, which is like the next chapter, right? So I just thought it’d be fun to talk about what is that and like, where does that sit in the ecosystem of tech?
Johannes Galatsanos (03:29)
Yeah, yeah. All right. We’ll start with the easy questions, Quantum computing for breakfast. So quantum computing is a field that emerged in the 80s and then more so in the 90s and then the 2000s evolved more. And it kind of came from a thought from Richard Feynman, who’s like one of the largest or biggest names in physics. And he
was looking at a problem of simulating nature. So nature is quantum mechanical. And quantum mechanical is basically quite counterintuitive to how we perceive the world, but it’s basically how very, very small things behave and very small and we say non-continuous things or continuous things evolve. Now, the…
What Feynman wanted to understand was how do molecules behave at the smallest level and if we could simulate molecules and could simulate physics and particles. So that will be interesting, for his perspective to understand physics better, understand the world and the universe and how everything is made, right, which is kind of core questions. But also he was trying to use it to simulate how larger molecules, for example, in your body, right, react, how do cells react with each other.
How do drugs react with your body? How does food get processed and how do you get energy and everything? So to really understand these molecules at the smallest, smallest level, said, okay, understood quantum mechanics, or we at least research quantum mechanics because nobody really understands quantum mechanics. But we researched that, we understood the basics of quantum mechanics. And now why don’t we use those principles to actually compute quantum mechanics with quantum mechanics?
And the main thought process is conventional computers. Even in the 80s, we had some basic computers right now. We have obviously much more powerful ones. But they scale exponentially in compute. So every about one and half years, they get twice as powerful, right? This kind of Moore’s law. So there’s a bit of a curve like this. But there are some problems where
The computation of just getting more more NVIDIA data centers in place and just scaling things doesn’t really help because let’s say when you simulate a little electron or like one atom, when you add another electron to the problem and you need a lot of electrons to simulate, you make each problem twice as hard.
So what he meant is that when you have a problem that you add, let’s say one more particle to a problem, right, to a big molecule, then your problem gets twice as hard. So let’s say you could simulate a small, let’s say, caffeine molecule, right, and you could see how caffeine interacts with your body, right, in water. Now, to make something slightly more
difficult, you know, you need, let’s say, two NVIDIA data centers, right? So you do something slightly more difficult, you need four NVIDIA data centers, and you need eight NVIDIA data centers, right? Then 16, and very quickly you’ve covered the entire earth and the universe at some point with NVIDIA data centers, and you can only simulate, you know, like aspirin, right, or ibuprofen or something. So that wasn’t a good solution. So we know classical computers will not help you solve these really fundamental problems.
you need a different approach. And that’s why he suggested using quantum computers, you can get an exponential increase. So one more what we call qubit, which is the fundamental bits in a normal computer, we have qubits in quantum. Adding one more qubit makes the entire system twice as powerful. So let’s say you have a thousand qubits, thousand one qubits is twice as powerful as a thousand. In conventional computers, thousand, you two thousand and four thousand and eight thousand.
So you have this exponential scaling and then he suggested, okay, let’s build that thing. Turned out that was pretty hard. So in the nineties, the first quantum computers were developed like toy problems actually here at MIT too. And then the first algorithms came out how to put these things together. And then over the last, I would say 30 years, but more so in the last 10 years or five to 10 years, actually, or not useful computing and not…
not useful quantum computers, at least reasonably powerful quantum computers came out and we could actually play with them and run some algorithms and show that they scale. So now there’s a pretty big industry. So there’s about a few hundred companies that have spun out that are trying to build quantum computers in various stages. And there’s also hundreds of groups across the world, like scientific groups that are probably thousands at this point.
that are researching both the hardware to build the quantum computer and the software algorithms around it. So as they develop new hardware and engineering concepts to build these more powerful computers, we can solve more more difficult problems. And to preface today, even though there are quantum computing companies that are already public, that have IPO’d,
Then one of them, for example, is Rigetti Quantum Computing. Interestingly, Chad Rigetti, the founder, was our first check in Diffraqtion. So it was also the first company that IPO’d a quantum computer to begin with. These quantum computers are still not useful for anything today, like an actual problem.
Your cell phone can actually simulate a quantum computer and run a quantum computing algorithm in there, but it will be much more powerful than today’s most powerful quantum computers. in a bit, so this is a question of timing. Some people say five years, some people say more like 15 years, so somewhere in between is probably the truth. These quantum computers will be more powerful than the best NVIDIA data center. And when that happens, then
The good thing is, you once you reach it, you just need to add one more qubit. You get twice as powerful as the best NVIDIA data center and then four times, eight times, 16 times, right? So you completely overcome the compute by exponential like a hockey stick, right? So, yeah, people are now betting on when is that, who’s going to win the race of getting first there, right? And which problems can we actually solve with it? And maybe to preface one last thing to know about quantum computers.
Well, there’s many things, but this one is probably the most relevant now in a five-minute summary. The quantum computers are special purpose tools, so they can only solve very specific problems. They’re not going to maybe run Zoom, like we’re recording right now, and they might not run Doc and run your Google Outlook and everything, but they will.
actually solve some very, very difficult problems that scale exponentially, which is chemistry and biology, pharmaceutical research, material research, things that simulate molecules like Feynman. Encryption is a big one. that’s all everything around why military and intelligence communities are interested in this part, which is cracking encryption.
making also encryption more secure. So it’s kind of a bit of a battle that’s going on there. And then there are also things like big optimization problems. Think airports or Amazon warehouses. You know, have thousands or tens of thousands of locations and you need to optimize things. So when you have very big problems, very difficult. So there’s kind of three areas where quantum computers will be useful and they actually are worth it to develop.
that’s in a nutshell the world of quantum computing.
Keith Cline (11:26)
Got it. Okay. Yeah. I mean, I just hear about it. And there was a company in Boston called Zapata that did their thing in quantum computing for a stretch, but yeah, it’s a fascinating area and it was just, you know, I guess it’s just only a matter of time. So, yeah. Well, let’s talk about your background story. So where’d you grow up? What were you like as a child?
Johannes Galatsanos (11:46)
Yeah, so I grew up in Greece in Crete, small island in the south. when I grew up there, it’s a small, particularly in a small village in the east of Crete. when you go out and you look up, actually, you will be surprised how bright the sky is at night because you can see the entire Milky Way. You can see all the stars and planets going on. So that got me interested from a very young age into also astronomy.
There’s even a small observatory up on the mountain where you can go up and see a few of the planets. So I was interested as a kid, you know, doing hobby astronomy. didn’t do it professionally because at some point I got also very interested in chess, or playing chess as a kid. And in 1998 or 1997, Gary Kasparov was defeated, which was the world’s grand chess master. I was defeated by IBM, Deep Blue, which is a chess playing AI.
And we were kids, I remember I was still there with my chess teacher and he said, maybe you guys should think about computers because chess might not be the best career move for you guys. So they kind of stuck with me and I got deeper into computers and AI and board games. So that drove me a bit into doing that more, tinkering more with coding, computers, software.
and going deeper into understanding how AI works and what’s the potential of AI, which also led to me then studying computer science. And it was in the 2000s, early 2000s, and wrote my thesis in AI. So took very few AI at the time. We only had a couple of AI classes in my grad school in Germany. But I took all the AI classes and wrote my thesis in AI, actually board games AI.
There was interestingly quite close to, so we had kind of defeated, know, beaten chess, right? Chess was solved, but Go was a big problem at the time. So Go was a much more complicated game because it has many more states and many more movements and positions. So we always thought, yeah, Go will be uncrackable. know, humans will always beat a computer in Go. So we’re looking at techniques of how to beat Go and…
A few years after, so I finished my grad school. I did a couple of years of research in AI. I left that field, but a few years after I left, AlphaGo came out. So there’s a nice movie, The Thinking Game, if you want to watch it, actually it’s pretty cool. I think it’s on Netflix. But there, they took it further and actually beat Go, which was a massive achievement. And that became AlphaFold, which was doing protein folding. Interestingly,
again, solving a big issue in biology, similar to quantum computing in a way, and that won a Nobel Prize just two years ago in chemistry, out of all things. So yeah, I know who would have thought. So there was a little bit of the initial trajectory of getting me into AI.
Keith Cline (14:52)
Very, very cool. So how did you get your career started after academia?
Johannes Galatsanos (14:57)
So academia was really interesting in the context that AI at the time wasn’t really being done in the industry, especially in Europe was a little bit behind in actually applying these things in the big companies. But at the other hand, it was in PhD in Germany and computer science at the time meant you’re kind of in your basement locked into and
programming some AI for five years and then you graduate. So that wasn’t the most appealing thing. But I actually wanted to apply some of these techniques to the real world and said, okay, let’s solve some actual problems now. And I got the opportunity to work with a company that had spun out of some McKinsey partners that were actually interested in building actual data and AI applications.
And they thought there was the future at the time. That was 2011 or so. they thought, hey, you we should really start building some cool AI applications and data applications. So this is the right time. yeah, I met them. They convinced me this is a really cool idea. Let’s start building applications for people. So I joined them and we started building applications in all kinds of things like energy, in like energy inspections with drones, kind of the first inspections like, in
thermal imaging of infrastructure inspections, know, like power lines that might overheat, right? And you could fly a drone and see if they, when they would overheat and need some maintenance or if trees were falling on your power line. Later, we did some automotive automations and big German automotive companies of how do you build, how do you automate assembly lines, right? And how do you optimize this assembly line robotic automation?
And yeah, there was super interesting work too. And then I got poached into Novartis, which was building, they just hired a new CEO that had said, you know, we want this big pharmaceutical traditional conglomerate to become a data AI company. And we were like, okay, cool. But nobody knows about data and AI in the whole company. So we had to build the whole skillset from scratch.
And yeah, I started there and stayed about seven years. And we built out the entire team for hiring data scientists, data engineers. And it was also a lot about, you can’t just throw AI when you have, let’s say documents, right? So you have, at least at the time, right? When you just had random documents lying around. I mean, I’m talking about physical documents, like printed somewhere. Like you can just throw AI and fix that thing, right? You first have to…
do a few steps till you connect all the systems, you build the data, and then you can run AI on top of the data. So most of my time was spent on digitizing all the information, making sure that they’re connected, we have the right infrastructure, we have the right ontologies and everything. And then on top of that, we could run different AI applications and suites for different customers. So
We built a pretty large team. When I left, there was about 200 people in that Data AI team, kind of serving multiple customers across the company. And yeah, was a fantastic time. It was a fantastic team. I really loved working there. But I did get in touch with quantum computing. So quantum computing, there were a few folks that…
came out at the time and said, hey, Johannes, I know you’re running all kinds of cool AI stuff. And I was always scouting for the latest and greatest AI tech that we could bring in. And then these quantum guys came and said, hey, you know what? You can actually now solve all these problems that you had, but you can now 10x, 100x, 1000x your applications. And I looked into it, and then I thought, you know what? First of all, I don’t get it. I don’t understand quantum enough to judge what’s going on.
But also it sounds like it’s kind of useless in what we do really, right? So it sounds like very far away from what we do. And third, like conceptually, and third, it seems like it’s still very weak. Like it’s still very early for me to really run this. So there was no real incentives to do proper quantum research or quantum computing work at the company. In my position, you know, if you’re like 200 people, you might get one or two PhDs playing around with quantum, right? So it’s a tiny, tiny part of your work.
And that would probably stay like that for 10 years. So I was like, okay, this is going to be at best a side gig for me for a long time. But it could be the main gig if I was on the other side actually building it. So I thought, all right, I’m not ready yet to build it because I don’t understand it yet. And I’m not sure if it’s actually useful enough to build it. So let me take a break.
let me go back to school and kind of focus on that and really understand it and see where it’s going, what are the actual applications and what does it really mean at the end of the day for a customer that is trying to use that. I thought MIT is the best place in the world to understand that, but also to understand about applications and how to spin a tech out like that. So I went to MIT, did the dual, did a
mid-career dual degree like MSC MBA, has focus on quantum tech, AI and applications of quantum tech, and then entrepreneurship. that’s what I did. And that’s where I met also then the rest of the founding team here, but I’ll maybe get to that a little bit later too.
Keith Cline (20:46)
Yeah, no, that’s a perfect segue. So let’s talk about Diffraqtion And I guess before we kind of get into how you met the team, just to to kind of know what you guys do, just kind of like to set the stage.
Johannes Galatsanos (20:58)
Yeah, yeah. So we built the world’s first quantum camera and the quantum camera is not exactly a quantum computer because, you know, at the time, you know, I realized, you know, quantum computing is still a bit out and you have to build this very large boxes.
Our quantum camera is actually quite small. I can show it here. I hope the filter picks it up. But basically this is the little quantum camera here, my little box.
So it fits basically kind of cell phone size. And the quantum camera can do a few things. So as my t-shirt says, you can see further and think faster. So it makes physical AI or machines and robot satellites, cars, both see further and then think faster. Now, seeing further means some, you take a picture with your phone, right, and then you zoom, zoom in, at some point it gets blurry, right? So it’s a blur resolution.
And you can get a better camera, right? And you can get better better cameras to kind of give you a little bit more resolution. But at some point you will notice even with the best camera, you reach you when you take a picture, you see a blur, right? And that blur is is a physical limit of what we call direct imaging or photography. Right. When you take an image, you can only see to a certain distance. Right. And that’s dependent on your lens fully, not really on your sensor, but your lens. Now.
If you want to see further, you have to take a bigger lens, right? That’s why people who do wildlife photography, they buy these big lenses, right? They can take or, if you went to a football game, you have people, these massive cameras, right? So because you want to see further to the action, what’s going on, so you need to buy these big lenses. Now, that’s a fundamental limit, right? So if you want to see further, you need a bigger lens. And now with the…
quantum camera, you can actually beat that limit and you can see things without increasing your lens. You can have the same lens, but see things which are further away. And that’s a massive deal because often you are forced to have like massive cameras, right, to see things which are far away, especially when you’re in space, right? Like obviously you’ve seen here the satellite floating. When that satellite wants to see something on earth,
It will see only very large things like buildings, right? Because it’s a small lens. And if you want to see smaller things than buildings, you want to see, for example, a car, right? Counting a car, you want to see a little boat, you want to see people, then you need a massive, massive lens, which is a massive telescope the size of, let’s say, the Hubble telescope, right? These are massive telescopes.
to these will cost you billions of dollars, right? So can only send up one, two maybe of these, right? It will take you 10 years to develop this thing. Now with the quantum camera, you basically have this. So this is the actual real size Hubble telescope. So this is actually the same power as the Hubble telescope. So you can see here the small lens it has. And this one costs you like a million bucks instead of a billion. So you can make it a thousand times cheaper by building something like this versus building
Actually, this is even a small satellite, but building a Hubble telescope. So the whole idea is, you know, miniaturize things like this, which actually need to see very far and need to also calculate things which are fast moving. And we do this quite the enablers, the quantum cameras, the satellite, we don’t build the actual satellite, we buy that. the thing that we buy, really and we can really build and are good at is the camera itself.
So yeah, that’s the main IP that’s side cuts invention. My co founder is a professor between MIT and UMD. And this really came out of over 10 years of research with NASA and DARPA. They were interested in finding exoplanets. So planets, basically life on exoplanets. So aliens. That’s why when I met Saikat, he told me he’s looking for aliens. I said, awesome. Let’s find aliens.
But also, what else could we do with a camera that can find aliens, So NASA was interested in that. DARPA was interested in what we call space domain awareness. So space domain awareness is you are tracking satellites in orbit that you want to keep safe from each other, right? They don’t collide. Think of it a bit like, you know, the FAA, but for satellites, right? So satellites don’t collide. You have some avoidance.
and you have little debris, right, maybe something that’s the size of this one, couldn’t hit your satellite and take it out because this one is flying at 20,000 miles an hour. So this is a hypersonic bullet. It will basically make a massive hole in your little satellite and completely destroy it. So you want to avoid these pieces of debris or space junk, right? So you want to track space junk, you want to track other satellites. Nowadays, satellites actually become more capable. They can actually move.
And bad actors can use their satellites, which are let’s say small and move them next to your satellites and do some damage to your satellite. It’s called like, you know, orbital warfare, right? That’s happening. So you want to keep track of all these things. So you want to keep your satellite safe. You want to make sure the other satellites don’t hit yours, right? And that’s called space domain awareness.
The problem is the satellites are very small, the debris is very small, they’re very far, right? They’re about 300 miles up in the Bofas. And to see them, you need a very good camera, right? And it ends up that you can only see things that is about the size of a cell phone. That’s kind of the smallest thing we can see with today’s technology in orbit. A cell phone definitely will take out your satellite, but even this can take out your satellite. So this one you need.
much more powerful cameras. So the lens is basically at its maximum. The limiting factor is the camera. So that’s what we built. And ultimately what this will be is a lot of cameras in satellites, in ground telescopes, and looking up basically to other satellites and eventually looking down to the earth, which is called earth imaging. So we will have a lot of these satellites looking down, up,
sideways and from up down. So that will be the constellation of things we’re building.
Keith Cline (27:26)
So that gets me into the next chapter of one, like the use case, right? So the commercialization of this technology. So you said, know, ground-based intelligence, so looking up. So that will be the initial kind of working with different contractors that are running satellite missions for private entities or government entities, that type of.
Johannes Galatsanos (27:50)
Yeah, correct. Yeah. So the cameras go into multiple satellites because there’s different use cases for you to do this, you know, traffic avoidance. If you think about like the FAA, right, so you have some ground towers, but you also have the planes themselves also have some avoidance, right? And they have pilots looking in all directions to see they don’t hit something, right? So in the same way, your satellite that you operate will have, you know, a camera to make sure they avoid any obstacles.
But also you will have ground-based and space-based applications looking and making sure that everything is where it’s supposed to be. And if something gets too close, you can make a little avoidance of it. You can get them a bit further. So the camera will be in other people’s satellites. Other people will buy the satellite to use that. And people’s telescopes, ground telescopes, you can put it in different places across the globe. And that data then becomes useful to end customers.
Keith Cline (28:48)
So how do you manufacture this? Like, okay, you’ve got the working prototypes or multiple. It’s like, then this is such a complex thing to manufacture. How do you even think about that next step?
Johannes Galatsanos (29:01)
Yeah, so there was a bit the daunting part at the beginning, but actually now it becomes from the beginning. I knew that it was also a mode in many ways of complexity, right? Of like building something like this. And honestly, also the main reason why this took this long to be developed in the sense of, you know, quantum computing theory started in the 1990s, right? And but
The quantum camera concept started only about 10 years ago. has been a very long time till people kind of realized this is possible. Now, that’s due to the complexity of what this entails. you need a few things. You need quantum information theory, which is basically understanding how the physical world works and how light works and how it can extract information from light. So, that’s one area you need to know, which is kind of algorithms and theory.
Then you need software, right? You need the software to run on top of the camera itself. Then you need data and AI because ultimately what you’re doing is you’re running AI at speed of light. So this part you need to have expertise in AI, which is kind of where I came from, And then you need people, of course, to build the hardware, right? So the hardware itself is, you know, you have optics and photonics, right? So you have
You know, telescopy, satellite, right? It’s there’s a lot of things coming together. And having such a unique skill set or let’s say having multiple unique skill sets combined into one to understand the applications and how we’re going to do this and pull this off. That’s kind of the tricky part. And it was funny when I started on this, I had met first Sai Khar, my co-founder, and then I asked him, okay, right, you’re doing information theory. I have the AI background and a bit of quantum.
we now need a hardware person going off optics, photonics, quantum optics. So we said, where are going to find that person? Because nobody’s ever built this. There’s not even a category. There’s not even a job that’s ever been described. So how am I even going to find this person that’s going to run this thing? And I went on LinkedIn. I just put in optics, photonics, quantum optics, 10 keywords. And literally, I kid you not, was maybe less than 10 people popped up from the LinkedIn search. I was like, my god.
Okay, so I got like nine shots to find my seat, know, right? That’s the only way to it. And I was like, okay, let’s start through the list. And then I noticed in the list, Christine, I said, hmm, wait a minute, we have, you she was a Harvard PhD in quantum optics. And then I saw the common context. One of it is a good friend of mine. So I asked her, hey, can you connect me to, do you know her first of all? What do you think? She said, oh, yes, fantastic, super smart. We were in the same Harvard lab.
You should definitely meet her. Let me do an intro. So I met her. She happened to live in Virginia and my other co-founder was in Maryland, so in College Park. So we told her, just, you know, cross the river. Let’s meet in on campus. Let me show you the tech, right, and the labs and everything. And then she came and said, wow, you can beat the Diffraqtion limits. Like I’ve done optics my entire life. I’ve never knew that you can do this. You know, all my life I could have done
all these other projects that I’ve done, I could have done that so much easier if I knew that this is possible. So this changes everything I mean. So basically, you know, we started working on it, putting together the DARPA proposal for getting some funding. got 1.5 million from DARPA for SPIR to kickstart the company. And then last summer we raised a pre-seed round and we get Chad Rigetti’s check too. And then we started building the lab and everything.
And even then we were like, okay, now we have three people. Along the way we met Mark. Mark Michael, he built Kepler, which is a space communications company. So he built satellites that are doing space comms and visual comms. So he knew everything about building satellites and that part of world. So, okay, now we had space. So we had that covered too. And then we said, okay, we still need another seven skills. Let’s hire people who have all of that. So it’s been an endless part of, which
thing are you building first, which skill you need first, right? And then buying the world leading expert in each of these areas to combine. So it’s been quite hard building something like this, but on the other hand, I know anybody who’s trying to replicate this, know, good luck. I try, you know, for us, it’s a massive challenge and we have the inventor of the tech.
and 10 of the world leading experts in this. Good luck if you want to replicate this. It’s going to take you quite a bit to build anything close to what we have.
Keith Cline (33:56)
Well, building in public is something some entrepreneurs might be afraid of, but you’ve been talking about what you’ve been building and there’s other podcasts out there of you. from what I think I heard was, wasn’t Mark Michaels like a judge or something? If you weren’t building in public, you wouldn’t have met him.
Johannes Galatsanos (34:17)
Yeah, yeah, that’s true. So, you know, we took that decision early on building in public because there’s a main constraint here. So there’s kind of two constraints. One that you need the world leading experts across multiple fields that are unrelated, somehow unrelated to each other. Right. So an optics person, like an optics engineer building a little
you know, camera system will not much talk to, let’s say, a quantum information theorist and then talk to an AI vision person and then talk to, you know, the world leading diffractive optics person, right? So these people are doing, you know, kind of similar fields, but they usually don’t talk to each other, right? So it’s not like any of us had all of these people in our networks because they’re just completely different networks. So if you want to
build something that’s really at the peak of innovation, you do need to find these people. And the best way to make it is you make the company attractive enough for people to know about it and come to you right when you put out a job description or that people know about it and say, these guys are doing diffractive optics. OK, I’m the world leading diffractive optics expert. I’m the world. You know, so get these people in. And the way to attract talent is to be out there, right? Go to
Go to scientific conferences go to technical conference going to space conferences to defense conferences, right and be out there So people know about you. That’s one part the second part was That Of course we’re building a novel tech like not just a novel technology a novel category, right? So this is a novel category of quantum cameras that people don’t know they don’t understand When you do this, you have to be kind of in people’s face. Like you have to say keep it
Hey, we have this massive camera, can do these amazing things. Look at it, look at it, look at it, look at it. And we have to keep repeating the message again and again and again again until people kind of get what it is and how it works. Because that’s creating that market to begin with, that market demand. And then thirdly, something I didn’t expect, but interestingly, when we published, we had a lot of contacts in Space Force, Defense, DARPA and so on, based on what we did.
when we started publishing our, when we did a press release and we announced things like in some of the defense journals, then we had a ton of inbound from people in the defense space, actually, and customers who reached out and said, Hey, you know what, I just read about this article. I really want to meet you guys. Tell me more. So we actually supercharged the BD so much. I hired a BD person because we had so much inbound interest that we couldn’t even manage it.
So we had the whole and now it’s the BD person and myself and my co-founder and we still cannot manage all the inbound because there’s so much inbound. People just love this stuff. we put the and you know, that’s by us being out there, right? And being public about it.
love for people to keep developing this tech because I think it’s a very, it has insane potential in multiple industries. So I’m sure this will be used in completely different ways that I’ve never even thought about it. And we will learn from other companies as they learn from us. So I much encourage people to build and compete and do more of this. B,
The main thing is that you will get is sure you will have a good camera, but the main thing that we’re building here is data, right? So data that you now see the world in a different way, right? You see the layers that you have never seen before in the quantum realm, right? If you want. So that part that you’ve never imaged before, you can see it in a new way because this camera actually doesn’t see in pixels and, you know, like this background pixels.
but it actually sees in something we call modes. So these are spatial modes. So this is a different way of seeing the world. And turns out if you look that way, instead of looking through pixels, you can now see further, right? So that’s the trick we do. And nobody has captured the whole world in this model world, like in this quantum world. So having more cameras out at as many use cases gives us data to build better models that we can use to run AI on these machines.
Really, the Edge AI use case is the important part and we build a mode through having as many cameras out there as possible, not through necessarily just being the ones that sell cameras. That was never the business model. So, yeah, we welcome most customers, right, which means you need to be public, best talent, again, public. Downside, you have competition. Not that much of a downside, honestly, for a completely new category.
probably more of an upside if there’s multiple companies attempting this, right? Then it means it’s a more serious tech that more people are interested in. So yeah, I think there wasn’t really a massive downside to being that public. And I normally encourage people, even in deep tech, know, go and share what you’re doing, you know, like, especially if you’re building something completely new, right? Then go share.
Keith Cline (39:38)
And you’ve been pretty scrappy. Like, so yeah, this is a deep tech company and a lot of deep tech companies require a lot of capital. when I look at, so correct me if I’m wrong, you’ve raised 4.2 million in dilutive and non dilutive funding and your most recent valuation was a hundred million plus. So that’s smart building by an entrepreneur. You know, it’s like you’re…
And you’re leveraging different programs. You did the engine blueprint, you did mass challenge, TNT, so these accelerators. So I just think other entrepreneurs need to learn from those lessons too of how to build a company yet in a capital and an industry that usually requires a lot of capital in a capital efficient way.
Johannes Galatsanos (40:25)
Yeah. So by now, this was in January, by the way, so we did rolling, what we call rolling admissions. And yeah, we’ve raised quite a bit more than that in the meantime. But something that was interesting for us to do was, you know, let’s say you get 10 million when you have it. Let’s say when you have a pre-seed.
company with a completely novel technology that has not been tested, that the market is unclear, the demand is unclear, it’s complicated, we don’t know where it’s going to end up, where it’s going to be really good and where it’s a less good than you thought. Then raising, let’s say, a massive amount instantly and diluting yourself massively and taking up all of that is probably a bad idea because sure, might get, let’s say,
10 million, right? And let’s say you get a 30 million valuation if you’re very lucky, right? In a pre-seed round, but most likely you’ll end up with something like, you four or five and then let’s say a 15 valuation, 10 valuation, something like that. So you’re already massively diluted on day one. And then you still need to, you know, how are you going to deploy this capital, right? You’re not going to spend five million dollars in three months, right? So you’re
Unless you buy super expensive hardware, of course, right, there’s different use cases. But even then, most likely, you’re not going to buy it all in the summer, right? You’re going to spread it out a bit. we built, I would say, not necessarily scrappy, but we build it in a way that is more organic, right? So we bought the stuff that we need at the time that we need. We hire the people that we need at the time that we need them. And we didn’t ever consider the budget in the sense of
Okay, we can only, let’s say we do need this person, but we don’t have the budget for that person. For that kind of scenario is exactly where I went back to, hey, okay, let’s do that progressive rolling admissions, right? So here’s a milestone we achieved, valuation went up, let’s get another check in and we buy this, right? Let’s go another check and buy that. So that’s how we kind of…
went organically up to this $100 million and a bit above threshold that we kept delivering milestones, kept delivering results, customer relationships, contracts. And with that, we could show, yes, the customer’s interested, the technology is progressing, the team is building. We get recognition and things like that, publications, scientific publications. And that builds this traction that allows you to go up to this $100 million valuation.
So I think it was a much more organic way for a technology like this, which is, you was very early, very, very niche, completely new category, right? So it gives you time to build off that without over diluting yourself from day one when you don’t even need that money. And ultimately we have raised more now, I think we will do an announcement soon, but we have raised more than we initially thought for the pre-seed or for the seed at that time.
So we raised kind of pre-sit and seat combined already with doing this kind of model of like, you know, incremental raising. So yeah, and that was honestly not the model we wanted at the beginning. like when we were raising, we also thought, you know, let’s just raise once and be done with it. But the truth is, you’re never done with it, You’re always fundraising. There’s always investors that will reach out.
You always have to consider that and build a relationship, right? So you’re not going to avoid that. think no founder, even though every founder hates it, nobody can avoid it. Everybody has to keep doing it. It’s a very natural thing. And B, yeah, you always need, you might need different amounts than you thought, right? When you start, especially a pre-seed. So that’s, I think two components that lead yourself to consider this kind of rolling admissions concept, right?
And lastly, it was so easy, right? The first check we came from Chad, you know, they gave us a couple of million and they basically just said, hey, look, you are in the middle of raising this massive DD round, right? It will take you months till you finalize all the contracts and everything. Now, you know, here’s a couple of million, go around, build something, just forget the fundraising for a month or two, right? Go and build it. Go back to fundraising after you have hired your first people, your office and everything. And it was great advice. You know, we took the check literally next day. We had the money in the bank.
Same day, I had put the job descriptions online and then we already had the first interview the day after. So on Tuesday, we had the first interview. So it was like, okay, bum, bum, bum, right? And otherwise we would have stayed another two months in fundraising. So it was actually a pretty good approach that I think more founders should consider for a pre-seed.
Keith Cline (45:21)
That’s a process. mean, I’ve done over 420 episodes and I don’t think anyone’s ever done it that way. The rolling funding process that you outlined. So I think that’s really interesting. And I had this sneaky suspicion. like, they’re going to have something new. Like there’s got to be another funding announcement coming soon. So I guess stay tuned. that what? Exciting. Okay, cool. raising for a new category, though, finding the right investors.
Johannes Galatsanos (45:41)
Yes, absolutely.
Keith Cline (45:50)
Like how did you perfect your pitch to get, you know, investors to write the check to fund your company?
Johannes Galatsanos (45:58)
Yeah, honestly, it was a ton of iteration and feedback. So, yeah, the problem was one, there is a lot of complexity with the different fields coming in, right? So the technology itself is complicated and then the use cases are very broad, right? So you have kind of two problems, right? And then third, people don’t know and have never heard this technology before, right? So there’s a lot of kind of credibility, technical, you know, validity of this.
And also how much, how useful will it be when everything is said and done, right? Because it’s early, you don’t know what the end goal really will look like, like what is the exact capability, the exact specs of this camera, let’s say two years down the line. It’s impossible to really predict, but you can make some assumptions, right? So there’s a lot of moving parts that also as you fundraise initially, the interesting part is, let’s say you have
started, you know, we just got the money from the DARPA government grant in, let’s say, May, right? So then from May to June, you have had one month, right, in the company, right? And then from June, from May to July, you already have 50 % more time or 100 % more time spent on the problem and actually spend on the R &D. So now you’re twice as smart, right? And also about the tech, about the customers and everything else. So let’s say every little
time that you get additionally in this initial phase gives you way more knowledge and experience, especially once you have some funding to build it. So your pitch gets better and more precise because you know your tech better, but you also get time to speak to your customers more and develop that relationship. So there’s one way how it naturally gets better. And then the other part really, you know, thinking like a VC does, right, that’s kind of for technologists, this is something that
Even when you run big budgets, big companies, you have your own P &L, it’s still quite different how VC would think in that particular time anyways, right? There’s a lot of thinking paradigms that change over time. And today’s thinking is different than even three months ago, let’s say before Iran, right? Or six months ago. So it’s important to have this constant feedback of like, what is interesting, what’s less interesting, and how to talk to, what element of that technology to…
bring up. So I think those three things made the pitch better and better, right? Getting constant feedback, talking with tons of people about it, and just distilling the most important part, And when you have technology, initially, you know, the pitch was probably way more technical. I think there’s a pitch of mine from a year ago or so online. And if you go to that pitch, there would be
10 times more technical than basically what I even said to you now. And that is a natural progression of you, like excited, very excited about the tech towards moving. you know what, actually these are the problems we solve and here’s just a tiny bit about the tech, right? So you can reverse that paradigm. But yeah, a lot of that is a nice balance to have, right? Until you get it right. But it just needs practice and iteration, yeah.
Keith Cline (49:08)
Okay, so you’re building this company in Boston. There’s a lot of great hard tech companies in Boston, but people might say, a space company though, that might not be like, so why Boston?
Johannes Galatsanos (49:21)
Yeah, we get the question a lot, which is in a way also kind of sad for thinking about, why does everybody ask the question? on the other hand, there’s a few very good reasons for us. One, all the founders have a big connection to the Boston area. So I was at MIT. My other co-founder was MIT, both professor and PhD, and then Harvard PhD from Christine. And we all have worked in the area, have contacts.
then there is quite a bit of, there was quite a bit of help, for example, the blueprints, the Mass Challenge program, a few of these programs, TNT and whatnot, that actually are in the area, and bring together brilliant minds. And these brilliant minds are a great talent pool of people who have gone to school here, they have maybe worked before somewhere else, right, even in space companies.
have done like, you know, a master’s or something or MBAs and then they want to they want to put this to work, but they do like the Northeast and, Boston living in Boston or living in Massachusetts, which is different than living anywhere else in the country really, too. Now, that is a great talent pool that you have direct access to, right? And kind of first bid, right? So you say, OK, you’re already here, right? Maybe you have a family or so or you just want to stay here. Fantastic.
You know, here’s a great company, you’re a world leading expert in XYZ. Come and join us, right? MIT is the best aerospace school in the country, right? So you get the best people coming from all over the world. It is really fantastic talent pool that you get first dibs, right? And that’s quite important, especially when you have these relationships and you’re known locally. Yeah, the second part, you know, I think talents and story was kind of the main part. And then lastly,
There are also quite a few customers because you and companies here that do the tangential things to what we do, right? Because again, it’s a new category. You won’t find a quantum company anywhere. So we can’t really hire people from somewhere else. But we have quantum companies, right? You mentioned, you know, there’s Quera, Atlantic Quantum, you mentioned Zapata, right? So you have quite a few quantum companies around that come spun out of the MIT Harvard Labs, which are probably the best in the country for quantum.
That’s a big community. You have photonics and optics, right? Great colleges. You have Draper Labs. You have quite a few companies around there on that. You have Defense. You know, have RTX, BBN and so on around here, Draper too. And lastly, you do have a little bit of space, but less. It’s said, unfortunately, not that much space in Boston. But while space is a big customer initially, it’s not our only customer.
The camera is, you your camera is everywhere. You have a phone, a new car, you have a new robot, and a drone everywhere. And these categories will increase and a lot of them are built here in Massachusetts. You know, we have Boston Dynamics, we have semiconductors, we have robotics. So all of that is quite central to Massachusetts too. And having access to that kind of both talent, but also customers is important. And lastly, manufacturing.
there are capabilities, there are some tax benefits. It’s more expensive than other places for sure, but ultimately getting the right talent, getting the right tax support and manufacturing support, and then being able to scale it here is also quite possible. yeah, we hope we can stay in Massachusetts while we go through the scaling manufacturing phase two, but yeah, we’re very committed to this place.
Keith Cline (53:06)
Yeah, I mean, it’s a no brainer for me. I mean, everything you said, talent, research, companies in the industry and visibility. Like your company is very visible here, whereas in other places it might be just another company that’s doing really cool stuff, but not getting the same level of attention and support. So I think it’s a smart decision. All right. So space in general is having this whole moment with Artemis and other initiatives. So what are you most excited about, you know, outside of your company about the space industry?
Johannes Galatsanos (53:36)
Yeah, man. Yeah, space has really exploded in so many ways. mean, obviously there’s the massive SpaceX IPO everybody’s kind of very excited about. So that’s going to be interesting. Also from a perspective of how will the launch cost go down, right? And that will enable also more industries, more satellites, more applications in space, of course, also reduce costs for our launchers, right? But that’s the smaller part. So all of that is super, super interesting.
B, think purely, you know, massive, massive Artemis fan. We’re like fanboying. Like I was that week of the Artemis mission. was two weeks. I was constantly, you know, going through, you know, the New York Times articles and my live feeds of like on Instagram and X and everything, what’s going on. So that was super exciting, super excited for Artemis three and then four, you know, actual moon landing, of course.
NASA has done some amazing strides too with now new leadership coming in with Isaacman and then they have quite aggressive plans for Artemis, but also general exploration, space exploration. And in that context, there’s unknown a bit, but there’s the Roman telescope going out. The Roman telescope is the successor to Hubble and the James Webb telescope. So there will be an optical telescope that’s also really tuned to looking for exoplanets. that will go, actually it’s scheduled in a couple of weeks or so. So yeah, stay tuned for that. Very cool science will come out of that. And then ultimately in, this is a bit further out, but like 10 years or so, the Habitable Worlds Observatory, which is the actual initial telescope that our camera was developed for with NASA, that is supposed to find a ton of exoplanets that are actually habitable, so not too cold, not too hot in the right zone. So yeah, obviously that will be the crown jewel for all space nerds in space exploration.
Keith Cline (55:43)
Well, Johannes, thanks so much for taking the time to walk us through your background story. Obviously, all the great work you and your team are up to in terms of what you’re building. to see what’s next.
Johannes Galatsanos (55:52)
Awesome. Hey, thanks so much, Keith, for having me and talk to you soon.