The following interview is a conversation we had with Jorge Torres, CEO and Co-Founder of MindsDB, on our podcast Category Visionaries. You can view the full episode here: $50 Million Raised to Enable Developers to Integrate Machine Learning into Applications

Jorge Torres
Thank you, Brett. It’s a pleasure to be here. 

Brett
No problem. So, to kick things off, could we just start with a quick summary of who you are and a bit more about your background? 

Jorge Torres
Yeah, of course. Well, my name is Jorge Torres. As you said, I’m the CEO of MySDB. I’ve been in technology for a long time, for as long as I can remember. I love computers, and every time I touch one, I still, or I marvel about how powerful this technology is. The fact that we can build something in our computers and provide value to hundreds or thousands or millions of people. And I think that has always been my motivation to build stuff for a long time. I’ve always dreamt about helping companies to leverage the power of artificial intelligence and machine learning to accomplish their goals. And now we live in these very interesting times where that is no longer science fiction. It’s actually very true. 

Jorge Torres
And it just happens to be that we’ve been doing this for a few years now, and we’re just kind of like in the middle of the best technology game that is being played at the moment, which is this applied artificial intelligence. My Co-Founder and I, we met in Australia over a decade ago, back at university. And even back then, we realized that this was going to be big. And we had a previous company that also was in the area of artificial intelligence, but it was too early and we basically failed fantastically. But we took all the learnings from that one and we started this one in early 2018. And, yeah, it’s been quite a journey so far. 

Brett
Was that company data CRM? 

Jorge Torres
No, that company was real life analytics. Data CRM still is alive. That’s a company I co-founded with a childhood friend of mine. And yeah, they’re still kicking. It’s just that it didn’t apply machine learning or it didn’t have any use cases for machine learning back then, and therefore, it wasn’t within my core motivations, but the company works well. Essentially, the nature of being entrepreneur is that once you do it once, you always want to try it again, and therefore, MySB is the third try it. 

Brett
And a few other questions we’d like to ask. And the goal here is really just to understand what makes you tick. First one, what Founder or CEO do you admire the most, and what do you admire about them? 

Jorge Torres
Yeah, well, I have many founders that I admire a lot. I think that given the nature of Mysp being open source, I admire fantastic open source founders. A company that I like a lot, confluent. Jay, I think he’s phenomenal. He’s done from the point of view of an observer, because once you’re internally, there might be a lot of buyers that you’re not aware of, but from the spectator point of view, they’ve done everything the right way. I think that Malgo, and the founders of Malgo, I think that there’s a lot of lessons that we still continue to learn from those companies. I admire, of course, entrepreneurs that face the general public opinion of what you’re doing is the wrong approach to solving a problem, and then they end up proving people wrong. And I think that, to me, is always fascinating at the end. 

Jorge Torres
Most startups are bad ideas until they’re not. And I think that what I try to do is not just to pick one single Founder or one single company. I try to look at those companies where MindsDB could be learning from, and I turn them into kind of like an aspiration for ourselves as a company and myself as a Co-Founder. 

Brett
And what about books? Are there any specific books that have had a major impact on you as a Founder and really just as a human? The way we like to frame that question is we ask for quake books. So it’s a phrase coined by Brian Holliday, and he defined a quake book as a book that rocks you to your core. It changes your worldview. Do any quickbooks come to mind for you? 

Jorge Torres
So I don’t know if quick to read, but certainly I love reading science fiction, because I think that science fiction inspires a lot of technology innovations, and then the other way around, technology innovations inspire science fiction. So there’s like this synergy between those two that I like, and there are four. Everything that is about science fiction and kind of futuristic ideas of how the world is going to play out. I love to read about those, because then internally, it feels like we can be participants of forging that future. And also we can use those as visionaries rather than just pure fiction. I like, of course, like most of us that are in this field, the foundation series. I like the culture series by Ian Banks. 

Jorge Torres
I think that those books are phenomenal, and especially in the place we are, where AI is evolving really quick. Those books were based on the idea that AI and any sentient being is part of a society. And again, intelligent, human like species, they are part of that society, but also AI. And we collaborate to accomplish goals that are grand. And I love that idea. 

Brett
It’s a perfect segue to dive deeper into AI. So from your perspective, being in this space and being a deeply involved expert, how do you describe the state of AI today? 

Jorge Torres
I describe it as kind of like the end of the AI dark ages, if you may. There was a lot of research in AI that took a long time for all the pieces to fit together. The compute needed to evolve, the frameworks on top of that, compute needed to evolve so that a lot of ideas that have been around for a long time could be applied. And then collaboration needed to evolve to the point of today where anything that any researcher, not researcher affiliated or not to a university, could share their output of their research, and then that output to be utilized by another researcher to even bring the frontier of understanding even further. Those things are only converging today, and they’re making the pace at which we understand innovation and artificial intelligence changes every day. 

Jorge Torres
I think that today you may learn one thing, and then tomorrow someone is going to publish something else that actually moves a frontier apartment you’re setting up. And given that there’s so many individuals collaborating to this task, we’re getting to the point where I is actually very useful for a lot of tasks that before were pretty much science fiction. The fact that computers now can understand language in many occasions better than humans can, it’s phenomenal, because language is such a fundamental tool for humans to build anything. I think that is probably the biggest augmentation of capabilities for us. So that’s where I see it today. The future. Still, there’s a lot of expectations and there’s a lot of unknowns, but as of today, I just think that it’s the most fascinating time to be alive. 

Jorge Torres
If you’re in technology, you can be participant of the improvements that we’re making to an objective of artificial intelligence. That objective is different for each individual. But, yeah, that’s the way I see it today. 

Brett
Are there any unknowns that you’re afraid of or you find particularly scary yeah, for sure. 

Jorge Torres
I think that pursuing general artificial intelligence, or artificial general intelligence, it’s a dangerous goal to pursue. I believe that maybe we shouldn’t be doing that. And I can elaborate more in that if you want, but, yeah, I’m afraid of that one there. 

Brett
Yeah, if you want to elaborate on that’d be awesome. I’d love to hear your perspective. 

Jorge Torres
Yeah, I think that understanding general artificial intelligence as artificial intelligence that should learn like humans do and better than humans do, and perform tasks that are human tasks better than humans do in all dimensions, is dangerous because it takes us in very tricky ways. For instance, humans are really good at being self aware, and again, there are humans that are better at this than others. But self awareness is a fantastic human skill. And if were to power computers with that, then how can you have a system that is meant to augment our capabilities when you actually, at the end, are going to have to convince these systems to do their job? One thing is to have very good AI that can augment you, and another one is to have AI that is aware of its position in its role society. 

Jorge Torres
And then that, in my opinion, trickles into autonomy again. Humans are also really good at becoming autonomous and self sufficient. And if that is the case, then we end up having systems that we’re going to have to convince to do the tasks that we need to do, and that’s not the goal. So I do think that there are human traits that we should never ever have on artificial intelligence systems, and they fall within the umbrella of AGI, for instance, survival, like humans, are phenomenal at surviving, and we have been doing it since the ice ages until today. If were to empower AI with surviving capabilities, then there’ll be a disaster. 

Jorge Torres
So I just truly believe that pursuing AGI, it’s a stupid endeavor, and we shouldn’t, but I think that we should pursue phenomenal artificial intelligence, and that is artificial intelligence that is bound by some rules that limit where we want AI to better than us and where we definitely don’t want to have any type of capability, such as survival, self awareness, et cetera. Super. 

Brett
Fascinating. Now, I’d love to switch gears here a bit and dive deeper into MindsDB. So can you just give us maybe the elevator pitch at a very high level? What do you do? 

Jorge Torres
Yeah, we help enterprises apply artificial intelligence into their systems, and we do this in a way that is democratic by the virtual being open source. We want to guarantee that no matter how big your organization is, you can still leverage on the top most outputs of AI for your business. And we do this initially by empowering developers to build AI applications in a way that is natural to them. I think that one number that is still striking for us is there’s about 30 million developers out there. There’s about 100,000 really good machinery engineers. And therefore, most companies cannot afford to have a top notch AI ML engineer. But what they can do is they can turn their existing engineers and software developers into ML engineers with the proper tooling, and that’s what we do for them. 

Jorge Torres
And again, all within the scope of applied machine learning, or applied AI, which is taking AI to pursue a goal that will make those companies be more effective at what they do and provide more value to their customers. So that’s what we do. 

Brett
And take me back to February 2018, when you were first launching the company. What were those early conversations like with your co-founders and investors and colleagues and friends? What were those early conversations like? And what was it about this problem that made you say, yes, that’s it, let’s go all in and solve it? 

Jorge Torres
Yeah, I think that at the time. So this is a good five, six years after deep learning became a thing. Deep learning is 2012, and were seeing how our understanding of machine learning was changing so much that the systems eventually will become so ubiquitous if there was a proper tooling for engineers to adopt them. So the gap that we saw was, okay, this technology is going to get exponentially better. And all we need to do is to guarantee that there’s a framework for people that are not in charge of the moving, the frontier of understanding of AI and machine learning. So they can just take whatever is the outcome of that and apply it. 

Jorge Torres
And I think that when we initially had these conversations with people, there’s always a division between believers and optimists, and then there’s always the ones that do things one way and are reluctant to change. So when we came to the conclusion that any developer could apply machine learning, some people thought, well, you know, that’s impossible because it takes a PhD for you to actually be able to do really good stuff with AI. But we also knew that researchers are not application developers, and the world is moved by application developers. So we knew that if there wasn’t a tool to connect those two worlds, enterprises were not going to be able to adopt this. So I think that early on we really saw that gap. I just think that it takes quite a while for those things to gain the momentum that we have gained. 

Jorge Torres
And I think that of those early conversations, the ones that were the most fruitful were are we going to open source MySP or not? And having made that decision from the very beginning to be an open source project, and then product has landed us where we are at the moment, where we’re one of the main players in the stage right now. 

Brett
And can you give us an idea of scale that you’re operating at today? So any numbers or metrics that you can share? Our audience loves to hear metrics. 

Jorge Torres
Yeah. So MySP, in terms of open source, has been deployed now more than 150,000 times. This is via Docker and Pip. Again, the product evolved so much that still we will continue to produce for the open source until the existence of MysP. But we also understand that we need to be a business. So we started a cloud version of MycB. And initially, this was a way for people to test MySP before they deployed it locally. So the demo cloud of MySP has about 13,000 users. And then, as of a few months ago, now a quarter, we started to provision dedicated instances for MySP, which we call MysP Pro. And we’re now at about 120 customers there. So now we understand better the funnel. We’ll continue to provide for the open source. 

Jorge Torres
They can deploy it in whatever shape and fashion they want, but at the end, the more sophisticated MysP becomes, the more complex it becomes to deploy it yourself. And there’s a lot of value for people to go and try it on a hosted version of Mysp. And we’re optimizing that funnel. We’re making sure that the experience that the first ten customers that we had, we can provide for many more. And right now, that’s a learning curve for us. How do we guarantee that the experience that someone had a few years ago when they went and deployed locally and they were happy with it, is the same experience that they’ll have five years from now, and that we can provide that experience to thousands of people, hundreds of thousands of people. 

Brett
This show is brought to you by Front Lines Media, a podcast production studio that helps B2B founders launch, manage, and grow their own podcast. Now, if you’re a Founder, you may be thinking, I don’t have time to host a podcast. I’ve got a company to build. Well, that’s exactly what we built our service to do. You show up and host, and we handle literally everything else. To set up a call to discuss launching your own podcast, visit frontlines.io podcast. Now back today’s episode. 

Brett
And what do you attribute to that success and the growth and the traction that you’re seeing? What do you think you’ve gotten right. 

Jorge Torres
I think that being surrounded by the right people at the end. Adam and I were co-founders of this, but we’ve been a startup all this time, and therefore, every individual that has joined MysP behaves pretty much like a Co-Founder, because there are so many hats that anyone has to wear at any given time. So Mysp success is a fruit of everyone that has been involved in MySp. And then when I say everyone is the core team of Mysp, which is small, we’re less than 30 people. And again, MySp could not be what Myspace today without them. Then there is the contributors that are in the open source. MySP could not be what Myspace is today without the incredible contribution of hundreds of people. We have more than 500 contributors out there. 

Jorge Torres
One of the cool things about MyCB is that virtually any place where you have data, MySB can go and get this data from. And the output of MySB’s machine learning capabilities can be put back into anywhere where you need that data to be. And most of those integrations are being built by community. So if you were to pin down one thing is people. The last part would be, of course, our different stakeholders, investors mindset. They all come with, course, money on the one hand, which is important to execute a strategy, but they come with experience, and an experience that is crucial, especially when a lot of what we do that is not core to the product, but core to the operations of the business, the go to market strategy, et cetera. 

Jorge Torres
Many of them have done this in the past, so we can learn from them. 

Brett
I think there’s a lot of talk out there about the benefits of going the open source route, but there’s not a lot of talk about the downsides. Can you share some of the downsides of being open source? 

Jorge Torres
I think that there are very few in the great scheme of things. Now, again, I drank this cool at 100%. So the answer that I’m going to give to you is as biased as it can get. If there isn’t one. Downside is that if you’re moving in a non open source project and you want to take a strategy where you take one or two customers, and then you guarantee that your product works exactly as those two customers want, then you don’t have to think about too many different variables. You just think of those two customers and whatever feedback that they give you, that is your product roadmap. In an open source project, you have an overwhelming number of variables, and understanding what is noise, what is not noise is very difficult. 

Jorge Torres
And then there’s a lot of lack of observability in the open source, just by virtue of the open source nature, where you do something, you send it out there, and then people will give you feedback in a way that is not the natural way that you will do it. If you just have one customer that calls you and tells you this is what doesn’t work, people do it either via GitHub issues. And it takes the right amount of combination of hate and love for you to put a GitHub issue there. You really have to love the product and hate it so much that you go and put the GitHub issue in hopes that someone will fix it for you. 

Jorge Torres
So I think that’s probably the only downside that I see, is that if you’re not developing in the open source, you have less variables to deal with. And therefore, if the people are giving you those variables are a good representation of your full market, then that may be a faster route. Super interesting. 

Brett
Now, talk to me a little bit about your market category. How do you think about your market category? Is it AI intelligence logic? Is it in database machine learning platform? What is that market category? 

Jorge Torres
Yeah, so the way that we see it and the way that we will see it can change a little bit, because right now the categories are being defined. But we understand that there’s enterprise data, and enterprise data is defined by the data category. You have like the databases, data warehouses, streaming, et cetera, and then you have the applications. And MySB understands that there needs to be a plot between those two pipe, if you may. So we call that pipe the logic, or the AI logic layer, and that’s where we play. We’re a pipe between the application and where people have data. And we live in this world where all these applications are now AI centric applications, or they will be AI centric applications. 

Jorge Torres
And the only way that you can turn them into AI centric applications is if your logic layer is also an app logic layer. That’s what we do. 

Brett
And as I mentioned there in the intro, you’ve raised $50 million to date. What have you learned about fundraising? Are there any lessons that you could share with founders listening in? 

Jorge Torres
Yeah, I think that we’ve learned a lot of things in the journey that may be tools that other founders can use, and there are other things that are specific to my speed, but I think that in the narrow ones, or in the generic type of advice that I can give is, if you’re an open source project, then there are investors that really understand how open source works, and therefore in the journey of fundraising, you should try to optimize for talking to investors that have invested in open source in the past, because they speak the same language, they understand the challenges, they understand where you’re supposed to be bringing investors that have not done open source in the past. We’re even trying to talk to investors, they will put you in a different framework. And sometimes it is really hard to ignore feedback from investors. 

Jorge Torres
And when the feedback comes from someone that hasn’t done open source in the past, they may take you in routes that are not the right ones. So, yes, advice number one is, depending on the nature of your software, if you’re producing software or any technology or platform, if it’s open source, then definitely there is a well established and very mature investment market for open source companies. Go and find them and meet them. And I think that’s one. I think that the second one is the value of accelerators and incubators. I believe that when you start a company, you have like this crazy amount of variables that you don’t know. And the true value of an incubator accelerator is to allow you to work on, figure out those variables with at least some capital that you can use to move your idea into a project. 

Jorge Torres
And I think that if I could advise to any entrepreneur, try to find one, be it y combinator, be it Skydeck, there’s a handful of those, but try to find one because they will give you enough of a tooling for the steps ahead, and also they’ll provide a lot of validation for investors on the line. And I think that the last one, for me, this concept of not all money is green only became evident as I started to face really difficult challenges in the company. And that’s when you really start to understand that an investor that is active makes a huge difference. Of course, money is super important, and money is a vehicle to accomplish your objectives. And if people are giving you money, then of course you should make the best exercise to understand from whom to take this money. 

Jorge Torres
But face onto investors, really try to. Or when you have to decide between a couple of investors, try to understand which one is going to give you the most value outside of that money. And that value will come in the most needed time. Because companies face so many different challenges and so many of those, as an entrepreneur, especially if you’re like an early entrepreneur, you will not know the answers until you have a brainstorming partner. And usually investors are fantastic at that, because they see so many variables. And if you have one that has done it before, then you’re making a decision that is a data driven decision. It’s not your data. It’s a data that collected. 

Brett
As you said, it’s really becoming clear right now. I was talking with a really good friend of mine the other week, and he’s raised like, 140,000,000 for his company. They were unicorn company, all that good stuff. And he said he goes to the board to try to talk through some of these problems, and they provide zero help. They agree it’s a very difficult environment right now. It’s a tough place to be. But he said that’s about as far as they get. And the board is essentially useless in the capacity to support him, help him navigate, help him build out of this. And I thought that was just very fascinating to hear, because I think the buzz in 2021 was, you just need cash. Cash is everything, and it doesn’t matter who the investor is, you don’t even. 

Brett
Want the investors involved. 

Brett
But it seems like now that’s being uncovered to be a mistake and a challenge that a lot of founders are dealing with. 

Jorge Torres
Yeah, we even faced this, like, on this last round, were getting offers at almost twice the valuation of what we ended up taking. And granted, our valuation was, I think, is perfect, but we made the decision based on, okay, what is the investor that we’re going to bring into the table that will make this company ten times better? And Mayfield, in this case, was the right candidate. Naveen has taken so many companies public. He was in hashicorp. So there are so many elements that we took into consideration that had nothing to do with the money at all. Because at the end, at some point, as you see, sometimes the markets become a little bit crazy. So there’s a lot of availability for money. 

Jorge Torres
But the thing that we want to avoid is what you’re describing to your friend, which is, okay, you take this money, and really, that money will get you somewhere. But challenges are going to face, guaranteed, there’s going to be challenges no matter what. And then it will be the people that you brought into the table that will say, okay, let me be a strategic partner to you. And that’s what we’re choosing at the moment. I’m not saying that we’re doing all the things perfect, but we also had similar experiences from friends of ours that face the same thing that your friend faced. And I think that right now, we’re trying to be very cautious about that. 

Jorge Torres
But the outcome of that is the advice that I was trying to solarize to your audience is, look, if the market is giving you money, which there are always steps in the markets because they fluctuate. Then when there’s a lot of money, just be extra careful on who you choose for your team. And luckily, we’ve gone through different cycles, but I think that in every single step, we’ve brought very involved investors, and that has made a huge difference for us. 

Brett
And were you tempted to take those higher valuations? How you describe it, you made it sound easy, but I have to imagine it’s somewhat tempting when you’re being offered these big valuations and a lot of cash. Is that hard to say no to? And if so, how do you work through that in your mind, to be more strategic and go with a valuation that you’re more comfortable with? 

Jorge Torres
I think that, of course it is really hard. It is really hard. But then it also makes it easier when you are not the only one making the decision. Like in this case, we already had a board of directors that was super involved, and I think that they helped us understand, okay, what is the right framework? And then for me, and this is for me alone, I don’t know how it was for the other shoemakers at that time, because at the end we wrote it, but for me, the bar was okay. All the people are giving us money. What are the ones that have been the most involved? Mayfield if we needed to talk to them Saturday night, 09:00 p.m. They will be there Sunday. They were there discussing the terms, kind of like ideating. So for me, those variables made a huge difference. 

Jorge Torres
If they were disengaged now, there is a probability that they will be down the line. And I think there were other funds that were like, hey, well, we can talk Monday or we can talk next week. And I think, okay, well, surely in the time of Cris of company solve these decisions, you need to go and work on the weekend and figure them out. And I think that if anything you can take from this conversation is that in every interaction that you have with an investor, you will get probably a good amount of signals as how will it be to work with them after they invest. And again, one thing is to have an investor that is of value, and that’s luckily what we found. 

Jorge Torres
Another thing is to have an investor that gives you money and they don’t add value other than the money. And then the last one is to have an investor that gives you money. And they’re also a pain in the butt. And that’s the worst. I cannot imagine what will be the situation for your friend if they couldn’t give him any advice? But also there were being a pain on their shoes. And I think that’s a one way to kill a company. So I believe that in every engagement that you have with investors, you will find variables that will help you assess this. It’s the same as when you bring friends to your circle of friends and things like that. So humans were trained to read these factors, and I think that investment shouldn’t be any exception of that. 

Brett
And final question, since I know we’re getting close on time, let’s zoom out three to five years into the future. Can you just paint a picture for us of that vision that you have for MindsDB? 

Jorge Torres
Yeah. So we know that every single application that exists today, and think about all of them, like email, calendar, CRM, ERP, internal tools, customer relationship tools, every SaaS tool that exists today will be reinvented with AI at its core. And we want to make sure that in the next year, every developer that is thinking of upgrading their infrastructure to an AI centric infrastructure, or any entrepreneur and their teams that are thinking of disrupting an existing technology with an AI centric approach, that MySP can be part of the stack at the end. We’re an infrastructure company, so our objective is to guarantee that we are the choice of developers that are building enterprise or the application with AI. And in five years, if you project that, then most of applications that exist today, we want MyCB to be part of the stack. 

Jorge Torres
Kind of like how today, if you think of MySQL, where they’re part of the stack of so many of the web 2.0 applications, they were the choice that most developers were choosing where they were building their web applications. I think that the journey of MyCB right now is to guarantee that the decisions that people are making today for what infrastructure they choose for their AI stack, that MyCB is one of those. And therefore projected down five years down the line, that we will be powering most of the AI tools that exist today. We’re not going to be the next email AI driven application. We’re going to be providing the infrastructure for those applications to be feasible. 

Brett
Amazing. All right, we are up on time, so we’ll have to wrap here. If any founders listening in want to follow along with your journey. As you continue to build and execute on this vision, where should they go? 

Jorge Torres
Slack. So we are super active in slack. And you can talk to not just me, my Co-Founder, our team, which again, as I was describing to you, everyone today at MySB, not only their shareholders of MySP, but also they act as co-founders and therefore anyone in the team will be the best person to reach over there. They’re super responsive and our goal is to maintain that for the foreseeable time. 

Brett
Amazing. Jor, thank you so much for taking the time to chat, talk about what you’re building, and share some of these valuable lessons and insights that you’ve learned along the way. I’ve really enjoyed our conversation. I know the audience is going to as well, so really appreciate you taking the time. 

Jorge Torres
Thank you, Brett. 

Brett
All right, keep in touch. 

Jorge Torres
Likewise. Pleasure being here. 

Brett
This episode of Category Visionaries is brought to you by Front Lines Media, Silicon Valley’s leading podcast production studio. 

Brett
If you’re a B2B Founder. 

Brett
Looking for help launching and growing your own podcast, visit frontlines.io podcast. And for the latest episode, search for Category Visionaries on your podcast platform of choice. Thanks for listening and we’ll catch you on the next episode.