Alberto Rizzoli, CEO & Co-Founder of V7

Tony Zayas 0:05
Everybody, welcome to another episode of the tech founders show. It’s Tony Zayas here, joined by Andy Halko. Andy, how you doing today?

Andy Halko 0:15
Fantastic, Tony, it’s another great day. I’m outside again, which I can’t complain about, but summer’s almost over. How you doing?

Tony Zayas 0:24
Good. Good, had a good Labor Day weekend to enjoy, you know, again, the last bits of summer that we have around these parts. So it’s been good. How about you?

Andy Halko 0:35
Yeah, things are great. So tell me a little bit. I know today’s guests. Really cool. Been looking through their website and and looking through everything. And I am actually I think the technology, so cool. So tell us who we’re talking to you today?

Tony Zayas 0:53
Yeah, we have an interesting one. We’ve had a number of founders on talking AI. And we have another one here. Today we have Alberto Rizzoli, he’s the co founder and CEO at V7. And V7 offers a complete toolkit for creating robust computer vision, AI, maintaining state of the art performance at every step. And so with that, I’m going to bring Alberto on. But before we do that, I just wanted to remind everyone, we have our upcoming tech throwdown, so you can see the site, for those of you founders out there who have a tech business, if you want to pitch, we’d love to have you apply, take a look at that page. And you can learn all about it. If not, you guys can just join later this month when we have the event and watch. We’re excited about it. We’re giving away some cash prizes, some services, it’s gonna be a lot of fun. So with that, let’s go ahead and bring Alberto on. Let me pull that ticker off the screen. Hey, Alberto, how you doing today?

Alberto Rizzoli 1:51
I’m doing well. Thank you for having me.

Tony Zayas 1:53
Yeah, for sure. So another international guests. We’ve had a few to this point. But you’re joining from London today. So that’s awesome. So welcome.

Alberto Rizzoli 2:03
All the way from across the pond. And yeah, thank you. It’s, it’s been a nice and warm week here. And it almost feels like almost as nice as where Andy seems to be right now.

Tony Zayas 2:16
Yeah, he’s got a prime spot there. Alberto, I, you know, I read the description of what V7 does, you might tell us a little bit more just elaborating.

Alberto Rizzoli 2:30
Indeed, we are a platform for transforming images into training data, which can then be used to turn itself into artificial intelligences. AI learns by taking training data, the image images and memories of the world that it sees around itself, and then turns that into an automatic program that is effectively able to to perceive the world much like humans do. On V7 teams can load images or video of say, a robot picking up apples, or a medical scanner, like a CT scanner, or a microscope that is analyzing a specific type of cell under under a particular condition. And then humans can go into the platform, and then instruct the AI what it should learn in a way that is very graphical by segmenting out a cell by creating a bounding box around an apple that needs to be picked up by a robotic arm. This is the process of labeling data permission learning. And what’s special with V7 is that there’s already an AI within it that can sort of decide cart with you, and label this data with you up until you’ve created enough that it takes off on its own. When you labeled enough apples on a tree or potential tumors in a CT scan, then it starts to take over and starts to make its own suggestion as to what can be labeled. And that helps teams that are building anything from autonomous cars, to robotics to next generation medical devices to effectively get to the to where they need to be with safe and reliable AI a lot faster, because they have a lot more of this training data to use in machine learning.

Tony Zayas 4:14
That’s fascinating stuff. So I wanted to talk a little bit about the AI. You know, this this concept of vision, you know, this vision, AI or computer vision. Do you mind explaining a bit of what that is? And how long is this been a thing? Because I find it interesting learning a bit about it that it goes it goes back quite a bit. And obviously you guys are doing a lot of incredible stuff with it now, but just to hear a little bit of that history, for those that are tuning in that might not may not be familiar.

Alberto Rizzoli 4:45
It goes back almost 80 years. In fact, the name V7 comes from one of the early experiments in computer vision, where some Swedish scientists discovered how the visual cortex works by putting electrodes in the brain of a cap. So the cool stuff that say neural link is doing or many other great startups in reading the the mind is also some pretty cool technology that is finally finding viability and so is computer vision. They come from the same branch of the dream of being able to create intelligence that interacts with silicon, or that is based in silicon. So in prodding a cat’s brain with electrodes, the Swedish scientists understood that the the visual cortex is actually made of layers. And that goes from understanding very primordial features like simple shapes and curves, all the way to things that we could almost consider memories like the face of the Mother, the cat’s mother or its owner. And this led to some real excitement in potentially using this form of graph based computing these deep neural networks that were just being experimented with a couple of decades later, to muddle the the the entire experience of the world. But that fell flat because we didn’t have the computational power to do so. So for many decades, we’ve created fantastic computer vision stuff by using what we now call classical computer vision, which is pretty much the concept of a computer vision engineer using mostly handwritten rules to define what something looks like. If you’ve ever gone bowling, or darts, you may or watch the tennis match, you may have seen automated detection systems that use classical computer vision to detect things. Also, many of the weapons of war that were created to to understand say the trajectory of something, or a heat seeking missiles were programmed by hand by a person. Today, though this is completely elapsed. And this is the cool thing about the deep learning revolution, that it’s moving away from a computer vision engineer having to handcraft these features into simply solving the task a few times, and then letting the computer figure out on its own how to program itself to solve it. However, until then, we could get to a whole lot of really great use cases, all the way from recognizing books, recognizing images on the internet, Google Search, Google image search was using a whole lot of functionalities. Without we’re using classical computer visions. It’s just that now that we can use these powerful deep neural networks, and we have the computing for that, we’re now on a rocket ship, there’s a lot more things that you can do, including stuff that, you know, would take forever, almost literally to perform by hand.

Andy Halko 7:34
Wow, I and I really want to talk about like the future of where that goes. But I kind of want to step back and learn about how you got into this. This seems like, you know, big initiative, a big, you know, product. How did you get started, where did this all come from?

Alberto Rizzoli 7:52
My co founder, and I had a previous startup to V7 called Aipoly that was also in the computer vision space. And it was 2015, one of the years in which this was just about to take off. And we created the first engine to run large, deep neural networks directly on mobile smartphones. And this was to help blind people recognize objects in real time. So we were using image classifiers, they would wait their phone around and the phone would speak to them by naming the objects that would see in his peripheral vision. And this was cool, although accessibility is a really hard market to break into. And we were finding ways of potentially monetizing this technology for good and in also a way that was monetarily sustainable. And we kept getting lots of inquiries from all kinds of business. It turns out that the things that people really care about in identifying that have some monetary or whatever we define as value behind them, are not always on the Oxford English Dictionary, they are a whole variety of objects that matter to you. Or that matter to your business. Say it’s a component of a car engine that you manufacture yourself. It’s a specific cell line. It’s a kidney but seen within a CT environment. It’s a car but seen from the camera of a self driving car from the rearview mirror using a very curved lens. So to create AI’s, that can work in all of these extremely useful business cases, you needed to create a set of training data for it to ingest and learn. And that’s where the idea for V7 came about, we realized that the for the world to we have this vision of creating a world with 2 billion programmers in it because when say AI really is working the way that we intended to anyone who can complete a task online, then even drawing something eventually becomes sort of a program of an AI that can can code itself. And to get to that point, we needed to make the solving of training data really, really simple. And once that solved, we believe that the neural network The texture side of things, which is what’s getting a lot of value today will eventually settle into a few paradigms or even something that that also solves itself out. And the we envision this feature of programming where you say, imitate a task, you say scrub your bathtub in front of a robot. And then the robot itself will see okay, I see are scrubbing the bathtub, I probably can do it. Okay, next time and it doesn’t attempt and, you know, gets there 70% of the way and then you do it again, yourself. And it’ll it’ll complete on its own. And, yeah, so V7 got started in 2018 in London been growing ever since we have we’re now in a period of very rapid growth as we’ve secured additional funding and are just on the way to what some people call Hyper Growth phases. Seems like there’s a hyper everything in the startup world. But it’s a it’s a fun time. And especially it’s a field that is booming. And it’s always good to be somewhere where we’re growth happens passively.

Andy Halko 11:08
Well, since we talked to so many founders through this show, you know, what did the early stage look like? You know, how did you go from this idea to actually turn it into a product and start to get customers.

Alberto Rizzoli 11:24
Yeah, so this is where it’s very fun. You know, when you talk to other when you talk to people in public, especially in public media or whatnot, you’re always talking about the successes, but it’s really the failures that are kind of the fun and juicy part of running a startup. And we’ve tried a lot of stuff to understand where I was eventually going to settle and be used in. And we’ve explored several, several verticals, many of which our own customers are now championing and taking to, to production. We’ve tried to build automated laboratory systems, we tried to build automated retail stores that you could just walk in, pick up a product and walk out at the same time as Amazon co was developing their first prototype. We’ve tried to develop a lot of feature product prototypes, and never seen the light of day, maybe one day, we will do like an expo ze assignment and Alberto’s failed products up until we we figured out how V7 should work. And the question that we kept asking ourselves is, no, after dabbling in a specific space, and meeting and interviewing some people that were in that industry, do we want to party with these people? Or do we want to have dinner with these people, it’s kind of maybe the more appropriate one. And eventually, we wanted to, we knew that every technology commoditize is and eventually becomes a little bit boring. But the people and their mission stay interesting for for many, many years. And we settled into the space that we’re in with our ideal customer persona being the lead of a machine learning team. Because we found them to be fantastic people to talk to, and they’re very much like Simon and I. And that kind of gives us this continuous level of energy, where every customer that we talked to, we kind of want to keep talking to them, and in a social context. So yeah, lots of lots of pivots along the way, in the early days before we had any money in our pockets. But then I think we found the niche that we truly love, and it takes a lot of effort.

Andy Halko 13:28
Yeah, if you don’t mind digging into that, you know, it takes a lot of effort. Was it an epiphany was it just that it came along? Was it strategic, because I’ve heard that before, too, is, I think, on this show, founders who, I mean, they kind of stumbled into that thing that was kind of the right fit and others, it was strategic, they sat down and they thought about the plan and they executed.

Alberto Rizzoli 13:52
It was several epiphanies. And I think that you need to have the discipline of a strategist but also the Mojo and the drive of a fanatic. To really find your, your correct space. It might vary from person to person, but I am personally someone who is very driven by my my drive to do something and very driven by inspiration at times. It’s not enough you also need to be in a market that works a to be very strategic about like, is this an industry that will grow and is also an industry that you know, has a measurable market size and has customers that we can reach out to but the other side was in it in a way it’s a feeling and it’s a feeling of pride in telling people what you do in the space that you’re in and and achieving that it’s a series of epiphanies and they try something and you know if it works out in the super early stages in which Simon and myself were really just on the whiteboard, coming up with ideas, we had the three day rule in which we would get really excited about something. And the mistake, we were making the mistake that many technical founders do in the early days, which is start with the technology and then find the problem to solve it for. And we would find these problems, we’re really excited about them. And then three days later, it’s almost magical, the 72 hour cycle, you stop being excited about them. They’re like, You know what, like, detecting, I don’t know floodings via satellite imagery, probably not my jam probably not went over, they wouldn’t be remembered for someone else out there. Who can be far more fanatical about it right? There from that space.

Andy Halko 15:41
I love that statement. And I just I want to sit on it for a second of that. You get you have an idea. And you have the sense of excitement, and waiting and get you know, giving it the time. I think that that’s that nugget of advice for other founders out there. It’s fantastic.

Alberto Rizzoli 15:59
Yeah, yeah, I would tell everybody, you know, let let it muster. It’s almost like you’re you’re you’re fermenting ideas in your head. And another thing is develop a lot of hunches. Because some of probably some of the best and most valuable ideas at V7 came out of multi month, or sometimes your multi year hunches where you’re like, Hmm, is this really like where the space is going? Could this be solvable? And then suddenly, you have an epiphany. So like, evolve them and keep them in your head. And V7, technically as as a concept in this kind of self evolving, platform for for self evolving AI that starts from the data perspective, from the training data perspective, is something that Simon and myself had an idea for back in 2015. Just at the time, we, we didn’t execute on it, we maybe we didn’t have as much belief in it. If we had started then we would be well, well larger than we are now. But you know, so it’s like.

Tony Zayas 17:04
So just to go back to that real quick. You mentioned the hunches, right? And kind of cataloging them and hanging on to them. What was the process that you want about you and Simon to examine those hunches? And then like, seek out to like, okay, let’s validate this to make sure that you know that we might be onto something or maybe not. How did you? How did you do that? How did you go about that?

Alberto Rizzoli 17:30
The validation part is some of the hardest because you naively think this is one reason why a lot of a lot of young founders going to b2c is because they kind of self validate with themselves. And instead, validating in a b2b company requires picking up the phone a lot requires embarrassing yourself in front of people that know their stuff, and you don’t. And we had to do a lot of that in the medical field, which V7 now operates as a market leader in the in the data space. We I’ve embarrassed myself so many times in front of doctors and like, I’m Italian. And I didn’t know the couldn’t think of the Italian word for drug in when talking to pharmaceutical companies in my own country, very early days. And I was like, oh, yeah, that the drug that determines pharmaco, but it just couldn’t, wouldn’t come to me. And, you know, you just look like this, this smug young founder, as many people will, will experience. And yet, you just have to keep embarrassing yourself, pick up the phone and talk to people. And eventually you’ll find someone who is extremely knowledgeable and patient that will tell you the lay of the land and discourage you to get into that field. And that’s when you know, there’s something out there. Because it means that the field is full of problems and full of things to solve. Yeah, so we we started by looking into problems, in fact, in the healthcare and life sciences side of things, and some of these are data points that mean nothing until you contextualize them by talking to someone. But the amount of medical data that the world produces is skyrocketing. And the types of devices that we’re creating to collect this data is even larger. Nonetheless, there is still there, there is a an enormous shortage of insight. There isn’t a shortage of people in a in medical AI. But there’s a shortage of insight because there are too many data points and devices for someone to fit in whatever the human version of RAM is their head. Doctors are fantastic at making a diagnosis when they meet the patient. But then collecting the whole patient journey into a system that can then present it in a in a way that contains suggestions to human is is truly valuable. And if you just ask an oncologist what the path is for someone who is going through through cancer therapy, it’s more stuff and more information points that any single person can remember for one patient alone, let alone a whole hospital’s worth. So that is one of the few areas in which you realize that the world of data needs a giant compressor need some orchestrator mind to at least suggest things. Then there’s other areas like the self driving car industry is creating enormous, they’re all under pressure to make cars that actually work and drive themselves. And so they, they have to, they have to create more training data they have to solve New Delhi, they have to solve snow environments, they have to solve what happens if a refrigerator falls off a truck and lands on the highway, the car is not recognized. It’s not trained to recognize refrigerators, what do we do? And this extrapolates to all other forms of robotics and automation. So we’re probably see in the next five decades, evolution of this probably five decades, the lifespan of this technology, maybe even less, but an evolution of many problems that we’ll have to solve before we evolve into whatever, whatever comes after artificial intelligence as we know it.

Andy Halko 21:08
So it sounds like in the beginning that, I mean, there’s a lot of different directions you could have gone. And we always talk about how do you narrow down and create focus and reduce some of the noise? For you guys, how did you, you know, that’s part of that epiphany and in what direction, but how did you keep your head straight of all the different ideas and directions you could go to say, this is where we really need to focus?

Alberto Rizzoli 21:40
Yeah, V7 as a SaaS platform, and like many SaaS platforms is one level upstream, and then a specific vertical case. So in the world of E commerce, we would be a stripe, where we’re an essential component of many businesses that need to execute within a vertical that they will truly love. Stripe, could easily build an E commerce platform that you know, sells anything, but obviously, they would they, they’re very good at creating this payment infrastructure for every internet company. And similarly, we create the training data infrastructure, for several AI companies that are all building their own device. And the need is as simple as you talk to in a speak in a field like us, which is full of newly emerging startups, you can talk to 10. And they will all be building something internally as a tool that is common across them. And that’s, that’s a billion dollar SaaS opportunity. And every AI team is building a labeling world was building now they’re mostly buying a labeling tool to get their raw data, just a bunch of images into label data, a bunch of images with demarcations of what’s interesting in that picture. And it’s sometimes as simple as that. You know, if you want to be in an emerging field, talk toward a startup founder and ask them what their internal tools that they’re building look like.

Andy Halko 23:16
So we did have a question that came in, you know, what’s the number one thing you know, now that you wish you would have known when you started?

Alberto Rizzoli 23:25
Um, there’s many. Let me settle into to one. And I always struggle with number one questions, because it makes me think like, I should say something very insightful. Now, I will probably just pick one of the first things that come to mind, which is how, how good humans are doing certain things. V7 is a platform that like tries to champion AI for the use of creating training data. And we’ve done some amazing work in there. But I feel like if early on in today’s we had a much bigger avenue to humans in the loop before turning things into neural networks that automate your task, then we would have probably gained like sort of six months on runway. This is something that many automation companies realize it’s sometimes actually the difference between AI companies that get a lot of funding and succeed. And the ones that die, is that some of them are very scrappy, and they will use humans to solve a task and call it an AI. And they’ll use a one component of AI. And as sneaky as it sounds, the AI will come it’s almost inevitable that many of these tasks will be solved. And so what you’re actually doing is you’re creating great market validation with something that is already an AI that is already pre built and also pre trained by education systems. So yeah, that that that’s one thing that comes to mind.

Andy Halko 24:59
Yeah, That’s great if you can, you know, fake AI company by just having people and calling them AI, you know, right. You’d be surprised. So, and I think that’s you, you mentioned what’s going to happen over the next five decades. AI is a big topic. I mean, I think some people don’t even understand it. What can you tell me a little bit about? Like, what are the big innovations happening generally in the industry now? And where do you see it going in the next couple of years, that we’re all going to be like, Wow, this is happening?

Alberto Rizzoli 25:31
Yeah, I think that, you know, in the space that both of us are in, in SaaS, it’s one of the more exciting fields because it’s AI is creating this third branch of software, I believe, we have backends and front ends. And now that’s this new data layer that is emerging. For privacy, you were creating data warehouses and creating kind of a database for a, a company, but the decision making on how to use the data was still left to humans, whereas now it’s left to autonomous systems that can, with one eye, look at the entire data that giant enterprise might have, and then extrapolate things from it. And I think that sounds both boring and incredibly exciting at the same time, because it’s like you’re taking the data from beginning to presidency, what gives you no, but it actually will have an enormous impact on like, some of the more boring uses of AI will have an enormous impact on on the world economy in ways that we don’t foresee, because we humans have this very simple eye, we have two eyes, and 10 digits, some of us. And this really slows down operating a business. If this is all the the importance of input and output that you have. Well as I think there will be many more tools for running segments or business autonomously, even just gently steering them that simply take data, and then take a broader knowledge of the world and the industry that you’re in and make certain suggestions, recommendations through things one way or another. Then there’s all the cool stuff, which is all the things that weren’t, were absolutely impossible before like self driving cars that are now doable with artificial intelligence. And there’s cases emerging every day. I think that even though AI still very much in buzz, we still have seen a very small number of kind of AI first companies being born. And we will, we will see many more of them. There’s there’s an opportunity for using AI and almost everything in this in this room. And this is something that almost every founder in a specific industry will tell you, before AI is for simple things like this, this cup was manufactured by a fired kelm, the downtime of Forge costs several 10s of 1000s of dollars a day or maybe even an hour, you can prevent it by having a camera or an inspection system that is automated. And the only way to actually have that is using it by using deep learning by showing it what a breakdown might look like once so it doesn’t happen again, or the onset of that. So there I think there’s a there’s a kind of a new new era of soccer with this data driven side of it. And then a new series of companies that automates small tasks in the world. And this will create this compounding effect in which you can run a business with far less attention given to micromanagement and far more to macro.

Tony Zayas 28:40
I, I have a feeling it’s gonna be along the lines of what you just shared there help are open, I would like to ask, you know, you mentioned a little bit earlier that human beings are really great at doing some things. And that’s an observation that, you know, you’ve gained in your role. How do you see AI impacting how humans work and perhaps, you know, evolving the way you know, where humans spend their time and focus?

Alberto Rizzoli 29:07
Yeah, there’s, there’s a big shift that technology is creating in automating tasks. And we know we’ve had the Industrial Revolution. And I think we can learn a lot from there. And now that we’re in a new industrial ocean, if you will, or information revolution, we should look out for creating the social fabric that enables simple tasks to be automated, sometimes even very complex tasks like legal or medical ones, while still preserving the concept of purpose and people. And I don’t mean jobs, because this is something that is in flux nowadays, as well as there. There’s far more independent workers than ever with the internet and especially of the millennial generation, anything after that for more people that just work for themselves producing content. And it’s a concept of creating purpose. And I think that we will eventually set that in a paradigm where AI will be very good at creating, and automating tests that are repeatable, and then it will become a tool for creating things that are new. And each person will have these set of tools to create something that is new and wonderful. For example, GaNS for creating new forms of art that you can have this paintbrush that can paint something that looks like a specific style, or we will eventually have forms of Photoshop where you could type in, I want a blue chair on the bottom left of this picture of a living room, and it can superimpose a perfect version of that will unleash this level of creativity. There’s a dilemma, of course, into what the what the future of humans look like, if AI can become very good at automating a whole lot of tests and doing on its own. And the answer of what society might look like is going to be for us to define, what we do know is that it is going to happen, because even if I were to disappear into a cloud of dust in a few seconds, the pace of progress of technology is something that humans have always been pursuing, and will continue to do so. And the Pandora’s box is open on this technology. So it’s going to be what, what does work look like when we have robotics that can do a whole lot of very tasks. And I think it’s going to be, I think we’re going to find out that there’s a lot more work for us to do on this planet than we first envisioned. And that in many cases, our jobs are simply what’s keeping society from collapsing, and to a large degree, and then a lot more of us are creating, you know, more benefits for society and growing it, there will be a lot more of the fulfilled, you know, society from collapsing maintenance of our world, making sure that you know, we have we survive, we’re healthy, we have food, and then work on things that are a lot more kind of higher on the hierarchy of needs. Now, I appreciate that I’m saying this from like, a privileged position living in central London, none of that. I do know that, you know, like, almost every industry ahead of us, there’s, there are big challenges, once there are big technologies that can really move mountains, in our horizon.

Andy Halko 32:32
Yeah, you know, and something that comes to my mind, and it doesn’t, I don’t know if it applies directly to you, but we want I talked to a founder, who had developed a really amazing, you know, facial recognition software, it was better than what what else was out there. So I’m kind of thinking of like, you know, what are the fears of privacy of, you know, data protection, and these other things that go into technologies like yours like this, that are, you know, scanning all types of systems and pulling in yet on, and you know, that we don’t, we may not even understand what really is being categorized out there and how it’s being categorized.

Alberto Rizzoli 33:13
Yeah, I’m glad you asked we, we have a stance against using this technology using AI in a number of fields. One of them is face recognition for the purpose of tracking or monitoring someone. Any form of racial discrimination, even if it’s in the onset of diversity, because you can really twist the story however you want. So, you know, detecting someone’s walking in a store, what background they have is, I think should be purposeless, and, you know, human intellect should be used for other purposes, primarily, and then anything that’s made to harm people, people, or in many cases, live animals that that shouldn’t be harmed. There, these are topics that we need to sit down and discuss with each customer. Before we accept them, we generally have a stance of no towards. And this is because we can, there’s plenty of business for us to go around. But that doesn’t mean that it won’t happen it will and there will be in there have already been several like without technology, there have been several deaths that have been caused by by AI that have been forms of surveillance that have been enabled by forms of human tracking in ways that the human did not give consent or to would not want to give consent. And part of this is a like the chant that the savior of for for us as a societal fabric for this willing part come from founders that speak out, and that will say like, we shouldn’t use technology due for this, but I will not change, I will not change the opinion of people in other communities where they believe that this is okay to do. Instead, it is the immune system of a country and a group of people that will need to react negatively towards the use of technologies to do potentially this this type of harm. You know, it’s like, the creators of the Manhattan Project couldn’t stop what happened with it. The the creators of any form of very powerful technology, the creators of self driving cars, will not be able to stop the fat technology leaking into military use cases, where you know, humans out of the way also means more, more and more impulsive use of tech. And so it’s a, it’s going to really be bringing this conversation to the world making sure that people are aware of the power of this tech, both for good and for, for negative purposes. Now, I’ve brought up a whole lot of topics in the introduction of the your founder, friend who is doing face recognition, which, on the other side, can save billions of dollars a year to an airport, by simply letting people through the gates faster. And every single person that would normally be late to a meeting, because spending one hour at the gate will no longer be it allows people to unlock their phone a lot faster, being able to unlock this thing, half a second faster. Monetarily makes an enormous difference to the planet. It’s it’s bizarre, but when you’re looking at this level of scale, it does make a difference. And this is some of the things that we need to put some of these technologies into perspective for that can be used for really some some really good things if they’re well controlled.

Tony Zayas 37:00
That’s all, you know, so fascinating to think about all the possibilities, from that ethical perspective, you know, what is being done within the industry, that you you’re aware of, to help provide guidelines and prevent misuse? Because that’s, obviously AI comes up, you know, to someone in the general public, and one of the first things that probably pops up his own, it’s gonna, the damage that it could do, right, even though we know there’s so much good that can come from it.

Alberto Rizzoli 37:32
Yeah, I think it’s one of those things where when you mentioned a lot of technologies that have one clear bad case, it becomes the champion in one of it. Another example is genetic engineering, people immediately start thinking of the trainings, they don’t start thinking of like an ALS. And that is simply, you know, part of how we formed our communication, forums. And pleasee I’m gonna do a quick switch. We’re running out of battery he says something was I can hear you okay.

Tony Zayas 38:19
It’s coming in

Alberto Rizzoli 38:25
now, is it better? Any better now or still bad?

Tony Zayas 38:43
There we go. That fixed it, the better. You’re good. Yeah.

Alberto Rizzoli 38:48
Okay. Sorry to everyone who was listening and got a robotic back some things of the things life. So do let me know if at any point I go robotic, or I glitch out for a bit of time, there’s stuff I can do. I missed a question here. What are the guidelines and what’s being done? So one really cool thing that’s being done is the whole field of Explainable AI is doing okay. But it’s growing. And that’s basically taking an AI and understanding why it made a specific decision. And there’s some great research happening in that space. There’s, I could say that there’s probably nine posited probably, for far more than nine positive use cases of AI happening or say, talent for founders that are doing something great. And for everyone that is doing something questionable. And for that use of it being questionable. The community it has a pretty strong immune system towards negative use of it. And thankfully, it is, we’re all standing on the shoulders of academic giants here and AI and all the tech that we’re using, and we’re glorifying it is really being researched, built and maintained by people who have decided to sometimes selflessly work for many years on on academic challenges that now lead our field. And when you go to a, an AI academic conference, which now numbers, it’s 6000 people, there used to be only a few 100. And now they’re super oversold. And, you know, people will be debating a lot on how to more ethically make use of this tech. And if you look at some of the giants in our field, say open AI, for example, has made it very clear that its language models should not be used for a specific number of topics that eventually influence their country in the United States. Voting habits. Natural Language Processing is fantastic for so many reasons. But, you know, it can be used as a weapon to to influence people to trick people. And so is so silly. So I think there’s this, this immune system response from the academic and research community towards companies that are making misuse of this, it can be stopped everywhere. And I don’t have a solution in mind for countries, especially where the individual rights of a person matter less, and therefore, the negative use of AI is more more doable. There is no remote switch for us to stop something like that, from being used. And there is no practical way of doing so. It’s going to then be a matter of public policy. And gosh, knows if we will have a public policy system that is strong enough to actually stop this use misuse of this tech globally. I don’t believe so. I think we just need to be very, yeah. Very, very open to debating it and making sure that individual people have. Am I robotic again? I’m hearing some static. Oh, sorry. Yeah, so we’ll know what misuse might look like from this tech.

Andy Halko 42:20
So I’m kind of curious, you know, you and your co founder, were you technical founders and built the product? Or did you, you know, have to bring other people on? How did you actually, you know, create what you have?

Alberto Rizzoli 42:34
Yeah. So Simon is truly a technical founder. He is an undergrad and Master’s in Computer Science. He started programming when he was six years old. So it doesn’t get more much more than that. And he, you know, he started programming in basic, so it was pretty hardcore. And he was kind of the, always the bread and butter of the technical backbone of what we built. I’m far less of a technical founder than him, I wouldn’t normally consider myself a technical founder. My degree is in business management. But eventually, when you’re enough in a field, especially in emerging fields, like like deep learning, you kind of become one because all the rules are new. But by far less, without someone like Simon, for example, I don’t think we could have built the first version of the V7 if even if we hired someone new. I think that, in some fields, it’s possible to I don’t think so it’s been proven many times, again, that to non technical founders can create a great business. But it certainly does help to have someone at least that can build a product from scratch. And I think it’s even better when two people build together. So Simon and I would work together on building the early versions of V7, I eventually took over. After we started hiring the every design aspect of the platform, I’m now handing that over. So I will probably not be writing a single line of code or making a single box on Figma in a few weeks time, but that’s just a natural progression. But until we got to that point, it’s a whole lot of work on building the thing.

Yeah, I think I’ve heard that over and over from people that we’ve talked to is having that, you know, key CTO or that technical founder. I mean, at this point, I you know, it’s I think it’s been the exception, not the rule that someone doesn’t have some sort of technical you know, person that’s part of the the founding.

Yeah, I think there’s, there’s even an adversity into venture capitalists funding your business if you don’t have a technical co founder. Which is sad in a way but you know, betting people need to make bets and they need to follow what’s more likely for them. So yeah, I would recommend if anyone listening is thinking of starting a new business, it’s always good. First of all, get a co founder, there will be days in which you’re feeling sick, or you can’t be bothered, and they’re the next best person other than yourself to look for solutions with. And if you aren’t technical yourself, if you can build a product end to end, then someone that can build a product end to end is going to be far more valuable to you than you know, someone that you you know, your friend or that is a brilliant marketer. Yeah, because if you do, yeah, everything will be decision that will be made at that stage will have ramifications for the rest of your company’s history. See Facebook being written in php

Andy Halko 45:52
Yeah, one of the other things that I’ve talked to a couple of founders about and I’m interested in is the, the mental emotional journey of being a founder. And I think a lot of times people don’t bring that up, or don’t share it. But it’s not easy. Sometimes you need somebody there to help you, you talk about the co founder, what’s been kind of your mental emotional journey so far?

Alberto Rizzoli 46:16
Yeah. Um, it’s funny, because I was just thinking so that the picture that is used for the ad in this in this podcast is this, this webinar episode was taken before I start right at the beginning of me starting V7, and three years later, I look 20 years older. And I think that’s just, that’s just the natural progression of things. You will, I do quite well under stress, and I think I thrive in it. But you if you don’t, you will have to learn. And you will have to surround yourself with other founders that you talk to even in auxiliar industries that have nothing to do with yours, simply to not feel alone. And that you will realize that half of your business success comes from your state. And, and that that’s that’s kind of important. That’s one other reason for having a co founder is that you can kind of support each other, you feel less alone, when there’s someone else who’s sharing the burden of like, every friend, because you tell all your friends, you tell your mom, I was like, Hey, I’m starting this. And then no one believes dual understands what you’re doing is neither overlay that you’re doing any of this pressure. So that’s super important. Find other founder friends, to have regular dinners with them. If you live somewhere that’s kind of isolated. And like you’re starting a company alone, I don’t know, move is my best recommendation or have regular zoom calls with them or whatnot. But move first, go stick. There’s reasons why people in the Bay Area are so successful in starting startups. And one of them is that every party you go to everyone is running or working as an early employee at a startup. And that just gives you that extra little bit of energy into doing things that differ the good from the extraordinary. And for us, it was it was a lot like this, it was like seeking support into each other, having other brilliant founder friends that were working on other AI startups or robotic startups or things that are completely different in biotech, and meeting them regularly. And, yeah, there’s other things you can do, you know, like, watch, watch. Y Combinator talks, even if, after you’ve, you’ve passed that phase just to reminisce that there’s a whole lot of other people that are embarking on the same boat with you, because the rewards at the other end are so exciting. And yes, I mean, eventually things all fall in line. And you might reach a state, which I believe is the one in which I’m in, in which now I become the emotional support of people that now need to start managing within V7 and Euro feel the pressure like, holy cow, how am I gonna, you know, lead this whole department of people, you know, in a field that is completely new, and it’s like a worked out, you know, the industry is booming. We just need to keep calm, build, build a rocket, build whatever we’re doing and keep going.

Tony Zayas 49:25
Alberto, you mentioned a little bit earlier, how you’re handing off I think design. And I would love to hear how your role as you guys have grown, how your role that you play as co founder? How has that changed? And where are you at now? What are you spending the majority of your time doing?

Alberto Rizzoli 49:44
Simon and myself are the kind of founders that would continue to build a product up until like we’re 500 people if we could. It’s very hard for us to lay down the pencil and manage and I think that makes some of the best managers are the people that really want to build stuff because they think they’re really good at it. And then you have to force them into managing others. And it’s a learning process. But I think that it is truly the indicator or like some of the some of the best people in the field because they know what good looks like. So I started by handling everything that had to do with basically the front end of the the business, from the design side of the things, the user experience, the marketing, sales. And today, like just even mentioning these names, there are several people in each and I couldn’t fathom, being able to do even even half of what’s being done in each in each sector. And as you grow, you need to start hiring people that are slightly better than you in each. And that’s the tricky thing is that when you’re interviewing them, they might not seem like they are because you’ve been, you know, when great marketing is at your company, you know, a great sales look like a company. So have people that are just wicked smart, and that like are are very all over into process or about mental flexibility and starting in a brand new scenario. And then building up from there coming up with new ideas to solve a problem. We create these like hypothetical company scenarios in which the interviewer and the interviewee are in the same level playing field. And then spend all of your effort in kind of handing over things to that person, even though you think that like you can do it slightly better, just don’t, because that person will become so much better than you by just fully loving and dedicated themselves to that to that sector of what’s ultimately your business isn’t the two of you. And it’s it’s a difficult thing to do because you you’re always doubtful as to like, should I be doing this should I be delegating? Start by delegating as soon as your your calendar starts looking gross. And that’s going to be very soon. If you’re doing things correctly. And then, you know, time, six weeks, and you you’ll realize like how did I survive without this? I have no idea.

Andy Halko 52:15
You mentioned that you are coming from an incubator today. Talk to us about resources. You talked about mentors a little bit, but just maybe dig a little bit more into that that is a common thread. We’ve heard everywhere. Use mentors. How have you used other tools like accelerators and incubators, or what else has given you that leg up to be successful?

Alberto Rizzoli 52:40
I mean, probably others will have mentioned it, but be greedy and get everything for free when you’re small. Because SaaS companies will give you free stuff, when they know they can make money out of you. That’ll just reduce the anxiety of runway just get Cloud Credits for free, get stripe for free, get everything for free, as much as you can. Yeah, perhaps not the best advice is in a in a SaaS discussion, SaaS got busy, but there’s always a chance that give it a try. And in terms of people, we’re lucky to have built a board of people that are very supportive and very useful for the company’s mission. I’m probably someone who should be reaching out to mentors more than I do. And in in our latest, latest funding, we’ve brought in some truly excellent people to access angels within the company that come from the academic community. But I think I’m, yeah, I’m not someone that spends a whole lot of time with mentors. But I will spend a whole lot of time with founders that have exited or that have you know, I hate asking for advice to these people, because I’m generally someone who, who, you know, feels very iffy about bothering people, but in a social context, you eventually start learning so much more from people that have done this journey before you and that eventually see themselves in you and vice versa. And so it’s a it’s an easy thing to pick up. So, you know, reaching out to other founders, especially those that have completed the journey or far more ahead of you is generally some some good advice. And then, yeah, watch out for for how you’re feeling mentally. A lot of people have burnout. Some people that may join accelerators, incubators, incubators, get imposter syndrome. Just don’t give attention to that stuff. A lot of people do but realistically, like your build, you have your own story. You’re building something completely different than other people. Don’t compare yourself to other startups in other fields, don’t compare yourself to people that are making millions and crypto. It’s not your story. And it’s not just not what you’re trying to solveocus on just yourself in your company and talk to others about like, the shit that they’ve been going, going into rather than like, you know, the keep away the ones that are continuously flexing about how well their their business is doing. And that they’re buoyed because they’re probably not telling you the truth.

Tony Zayas 55:37
That’s pretty good. Alberto, we have a few more minutes. But I would love to hear before we ask our last couple of questions, what’s on the horizon for V7, in the next year.

Alberto Rizzoli 55:52
So much stuff, it’s a pretty exciting time, because we now have so many talented people that have a vision for every little bit of what we do. And so it’s almost like, I log on to Slack every day. And there’s like a new feature that like I haven’t even monitored in the beginning. Okay, who designed this, we built this. And that’s one of the coolest parts right now. But we to keep it brief, we are moving the user experience of the platform to what we believe is the next stage of what we’re training data platforms will look like. And we’ve started by making a product that’s very, very highly technical. And that’s going to help the most pro our users, the ones that like really need a very scalable way of creating training and managing datasets are enormous, and then shoving them all into compute to train the models for say, a self driving car or a medical scanner. And we now have the capability and bandwidth of creating an experience like that that is ever more magical and ever more self alternating. In practical terms, what we’d like to do a year from now is create a framework to solve any task that requires site and decision making through the V7 platform and just through the graphical user interface. And yeah, so that that’s kind of what what the the engineering and r&d is putting their heads around. And we’ll see it’s, it’s going to feel like a whole new software building experience. I hope and, and, yeah, that’ll that’ll be ready in the in the next few months.

Tony Zayas 57:41
That’s awesome, super exciting stuff, I would like to say we usually extend the invite to our founders to come back when you have an exciting update. So with all that, you know, on your plate, we would we would love to have you back to share more at some point in time. But until then, where can our viewers learn more about V7, follow V7 and you just to see, you know what you guys are up to and participate.

Alberto Rizzoli 58:08
So if you’re a founder that wants to build a computer vision company, or using computer vision, or any form of imagery that needs to be classified in any way, best way to try out our tool V7labs.com That’s letter V number 7 labs.com Just Google V7. And just press the big Get Started button. And that’s that’s how you get access. And the Twitter handle is also V7 labs as well as LinkedIn. That’s where we post some updates. And you can also reach out to me directly via LinkedIn via Twitter and DMS, especially if you’re if you’re in this field, or if you want to talk about startups or anything like that. And I would absolutely love to come back on the show. Maybe we could do a retrospective maybe because it like Alberto from September 2021 was so stupid, as it usually is like every every year and kind of feel like I was such a dumbass a year ago. So hopefully it’ll be a continuing.

Andy Halko 59:09
That’s great. So going right off of that. My final question that I always ask is, if you were able to go back in time before you started the business and have coffee with yourself, what advice would you give? Hmm.

Alberto Rizzoli 59:28
As soon as there’s investor interests in your company, spend half of your time if not more looking for and hiring credible people. Because it means you’ve settled on an incredible idea or product. And you and your co founder are not enough to turn that vision into reality. So yeah, just find the most incredible people in the world. And when when there’s fewer of you, ironically, it’s easier to hire some some fantastic hires than when you’re in like 20 headcount or so because first of all, you’re bogged down by, by so many people that like finding that unicorn employee becomes a bit harder. And also you you’re just, you know, on this mission, he says, let’s, let’s scale this up and let’s build it. So that that’s probably it. If you do if you are kind of in in that stage right now, as listening and yeah, just the key is whenever you meet an incredible person that you know, at a hackathon or whatnot, that’s how my co founder I met, keep them in the back of your mind. You never know when when they they might end up working with you.

Andy Halko 1:00:47
I think that’s fantastic. Thank you so much.

Alberto Rizzoli 1:00:51
Thank you both. Thanks for having me. And, yeah, lovely to chat for the past hour.

Tony Zayas 1:00:56
For sure, Alberto, thank you so much. Alberto, Rizzoli from V7. had heard you guys go check out his site and just you know, see what they’re up to really fascinating stuff. So thanks again, Alberto. We’re appreciate your time. This has been outstanding to everyone tuning in. Thank you guys, and we will see you again next time. Take care everybody.