Jorge Torres is co-founder of MindSDB, a tech company that's working to democratize machine learning so businesses make better predictions with their own data. He speaks candidly with Elian and Juleyka about pushing himself to think bigger and reimagine the problems his company can solve, finding the elusive product-market fit, and learning to navigate the VC funding game while rising above industry bias. Our Latino Business Moment of Zen affirms our significance and the value we bring to the sectors we operate in.
This episode was supported by our LBAN Alumni Founding Sponsors Windrose Vision, VND, Fig Factor Media, Allegiant Electric and Global 4PL.
Elian Savodivker:
Hello, and welcome to Scaled, the Latino Business Story. I'm Elian Savodivker, director of engagement at LBAN, the Latino Business Action Network.
Juleyka Lantigua:
And I'm Juleyka Lantigua, LBAN alum and founder of LWC Studios.
Savodivker: On this show, we speak with Latino business leaders who have grown their companies to one million dollars or more in annual revenue. Our guests share their scaling stories and business insights with us and help us explore the world class research coming out of LBAN.
Lantigua: Our guest today is Jorge Torres from MindsDB, a tech company that uses machine learning to help businesses use their data to predict the future. I was so happy to meet Jorge because he really helped me to understand how machine learning is changing everything in his sector, and he also gave us such great insight about how he's been navigating Silicon Valley to make his vision come to life.
Savodivker: It's so inspiring to see leading Latinos in these industries, and I loved what he had to say about the importance of finding the right product market fit, something critical in a startup, how he's constantly pushing himself to think bigger and look for a larger problem, the bigger picture, right? That's part of his strategy.
Lantigua: The best part was that even though he shared so much, I left wanting to know so much more. I had so many more questions. I wanted to keep learning from him, and that's very exciting for someone who is in a completely different field, to have someone who is so impassioned by what they do that they basically make you a believer. Honestly, I'm so excited for him. I'm excited for MindsDB. I'm excited for Latinos in the sector. I think we've just started to understand the impact that we can have.
Savodivker: Well, Juleyka, there's so much groundbreaking work that Latinos are doing as entrepreneurs. We're going to have a next season, right?
Lantigua: We better. We're just getting started telling all these incredible stories.
Savodivker: So with that, here's our conversation with Jorge Torres.
Torres: My name is Jorge Torres. I'm the CEO and co-founder at MindsDB. And at MindsDB, we make it very easy for people to apply machine learning to predict the future. Essentially, if you have data, you try to make decisions based on intuitions that you have from that data, but MindsDB helps you use as much data as you can to get a better prediction of what's going to happen, either in sales, inventory, costs, et cetera.
Lantigua: Before I even get to my first question, I need a little vocabulary lesson. So Jorge, can you help me understand what you mean by machine learning?
Torres: Machine learning is the science, if you may, of teaching a computer how to solve a problem without human input at some point. So imagine that you want to label text that may be good or bad for your reviews on your retail store. Traditionally, a human will have to read this text and then say, well, actually, this person is thinking that the review is positive or negative, or that the product is amazing or the product sucks.
So what you can do is you can code an algorithm that instead of having rules that says, okay, if someone says the word I hate or I love, but more like read the whole text and then on your own come up with some rules to identify if this is positive or negative. So machine learning does exactly that. It's that the techniques and mechanics that engineers use so that they don't have to code every specific rule for a specific problem, but then you can give it some examples. And then based on those examples, the machine can learn itself how to do it for data in the future that doesn't have the actual answer. And it'll give you an approximation of what that answer can be.
Lantigua: Wow. I am definitely not a technical person, but I am a complete science geek. And the more I learn about AI and automation, those kinds of things, the more I get into it. But you guys, you, and by you guys, I mean Adam Carrigan and you, your co-founder, you guys have this notion of democratizing machine learning. What does that mean, and how did you decide this is what we want to do, this is what we want to figure out?
Torres: So what we learned is that people don't necessarily need to know even what machine learning is to apply it and to make it useful, just as when you switch the light in your house on, you don't have to know how electricity works and how that even got to your house. You just know that you're using it. And you're using technology every day. You start your car, there's a whole bunch of engineering behind it, and you start your phone and you don't even have to know how internet works to use it.
So for us, it's important that machine learning gets to that stage, that you can ubiquitously apply machine learning to make better decisions without really knowing that you are. And at the end, we believe that MindsDB will get to the point where it's more about having a conversation with machines, and then you get the answers that you want to have.
Now, the journey to get there is a long one. So we do it by stages, we do it by tiers of understanding. So the first tier is all these developers that are not necessarily machine learning engineers, but are closer to the technical problem and they can apply it. And that's what we solve in kind of the first section of the journey of MindsDB, with our open source project. And the second part is, okay, in the commercialization stage, let's make sure that we can provide the same capability for people that can ask questions to a database and enable that of being able to apply machine learning to get questions that are predictive.
And we believe that the next step is, okay, let's make sure that people can ask these questions without having to write questions in a specified query language, but they can do it in natural language. And we believe that when we accomplish those three stages, then we will fully have democratized machine learning for a lot of use cases. Does that make sense?
Lantigua: It makes a lot of sense, but I'm also thinking anthropologically, you also have to teach the machines how to speak human, in a way.
Torres: Yeah. It is hard, but it is a problem that is improving exponentially. So what you had five years ago, it's nothing in comparison to what it is today. Today we have systems that will fool 70% of the humans that will have a conversation with those systems. And five years ago it was 10 to 20%, and in the next five years, it'll probably be 90, 90 something percent of the humans that will have conversations with the systems.
And even if they know that there is a machine behind it, they're getting meaningful answers and the conversation is fluent enough that you can still extract the information that you want to extract in a natural way. And I think that in 10, 20 years, maybe no one will be able to tell the difference. These machines will be able to be as useful at solving some tasks, to the point that then the human doing those tasks can be doing something more interesting, like making sure that they can leverage the work of machines so they can do at a bigger scale or focus on even harder problems.
Savodivker: I want to jump in here, and I want to talk about this process because obviously this is a long journey to get to where you are today, and that takes capital. And I want to talk a little bit about that journey. We know from our research that our 2021 report, for example, talks about how Black and Latino owned companies have obtained only 2.6% of the total VC funding that's been allocated. Now, MindsDB has raised over seven million dollars in seed funding. Can you talk to us about that journey? What were the challenges, and what were you able to really learn from that?
Torres: Yeah, of course. I think that the biggest challenge was to understand that it's a game. Like any other game, has its rules. And if you focus first on playing the game without really understanding the rules, then the game will be over by the time that you've learned the rules. So once I really understood that here in the Valley, people have been doing risk investment for such a long time that the rules for different stages of funding are very well defined. Like, okay, you want to start a business, you don't even need to have the capital to do it. You can do it with an idea. You just need to talk to special type of people that have that appetite for risk.
And then once you do that, then you have certain milestones that you need to accomplish to get to the next stage. And then that comes with a whole other set of rules of who are the people that you're talking to, what is the shape in which you deliver the message of what you have accomplished? And then what are the milestones that you're going to have to accomplish to the next one? And even now, even though we've raised that amount of money, we're still playing the game.
So we've accomplished our milestones, and now we're also in the other stage of now we need to raise a bigger round, and it has different rules and we have to learn what that is. So I believe that coming from Latin America, where if you want to start a business, you have to really have the money up front, and day zero, you have to make sure that that business is making money. A sustainable business is something that needs to happen right away.
And that mindset is not necessarily how businesses are done in the environment of risk investments. And then changing that mindset has helped a lot, because then you can think bigger, you can think as well, okay, you don't need to solve a problem for today of profitability. That is a problem that needs to be solved, but first you need to solve the problem, okay, are you addressing a problem that is very large and is a problem that can really have a much larger impact?
And I believe that those stages and that framework of thinking is what's very unique of places like this one, where people can think of crazy, crazy problems and crazy, crazy solutions to those problems, and give it a crack. There's going to be people that will back you into that vision.
Savodivker: So I want to talk a little bit about that because you're a Latino founder, you come into Silicon Valley, and you have to play this game. For those that are listening and that are maybe going through that initial stage, what was that game like? What did you have to play as a Latino founder to get the funding that you needed?
Torres: I think that we were very lucky, and I call it luck because by pure chance we ended up in an incubator. And I think that the purpose of an incubator is really to get you into that mindset, really help you understand that maybe your ideas of how this works, and how it actually works, may be off. And then it's a period in which you can train your assumptions as to what you should be doing at a given stage. So that helped a lot.
And I think that the best advice I can give is to try to get into an incubator because incubators will give you that mindset. And also, incubators have this mindset of trying to be inclusive. I think that the amount of money that has been delivered to Latinos in venture capital is very low. That is changing, and is changing very much at a rate that is very impressive. But actually, incubators tend to be super open about emerging markets and people that come from all over the place, at least in my experience. And therefore, that should be the first place to start. I think that the problem is that people don't apply because they're not aware of that being step zero into the game.
Savodivker: I think that's really powerful advice, and definitely to be on the lookout for those. Now, you talked about the game kind of changes, different stages. You're at a different place now, and Forbes just recently recognized your company as one of America's most promising artificial intelligence companies. You're going into some of the bigger stages for funding, series A and above. As you go into those stages, what do you think needs to change as you play the game, or what are going to be the different challenges that you're going to be facing?
Torres: Yeah, so there is this crazy learning that I've had over the past few months, which is even though the rules have been set for a long time, at the end, people have interest that they have to take care of. And depending on the economic situation of the world, then they have leverage or not. So let me go back into what I believe so far is the learnings that I've had over the stages. I think that when you're idea stage, you're in the angel stage at this point, a PowerPoint presentation. You just have the idea. More than the idea is you have a problem that you love. I think that many entrepreneurs, we start with, oh, I have this product and this vision of a product. But I think that it's more important to think of, what is really a big problem that is not being solved, and is this a problem that a lot of people have?
And then being able to convey that that problem is huge. And if you can really fall in love with a problem, then you will find angels that also feel that that is a problem that would like to contribute. Many of the angels, they're just doing it for being involved in this kind of lines of view that aligned with the way that they see the future. At this point, thinking of what the solution is, is there something that you can do right after you get that first check from an angel? I think that now you have capital to start thinking, okay, what would be the solution that I will have?
And once you have a prototype of that solution, then I think that is when you get into the seed stage round, when you're like, okay, well this is the massive problem that I have, and I think of all the research that I've done, this solution can work. And there will be some people that will buy into that combination, essentially product and a market, but not necessarily that there's a product market fit. They believe that you have a vision and they believe that where you're starting can get you to a point where your product will eventually iterate.
And they all know that at that stage, your product will change. What you're telling them may be different as you collect information, but you've shown and convinced someone that this is a good start. And I think that that is a seed stage.
Now for Series A, people gave you money to get to the point where your product changed enough that eventually you find what is known as product market fit, which is your product and the problem are extremely aligned and people are willing to pay for it. And that's where we are right now.
Lantigua: All right, let me pick up on that because I love a thing that you said, which is pick a problem that you love. I think that that is such a beautiful way to encapsulate what drives an entrepreneur. I mean, we are stubborn so-and-sos, but it's because we love the problem that we've picked to focus on. But finding the market or defining the market for the solution to that problem can sometimes be one of the hardest challenges that we face. So how did you and your co-founder define the market for the solution you were building for this problem that you love?
Torres: Yeah, that is a great question. And the candid answer is that we fell in love with the solution before we fell in love with a problem. Luckily enough, we fell in love with a solution that was behind a problem that was very large. And then we found our ways to fall in love with the problem itself. I think that LBANitself was one of those enlightened moments where “okay, maybe we're seeing this wrong.” We're seeing this from like, okay, this is the product and there's these people that we know that have kind of that need.
But I think that when I did the program, it just really changed the mind as “Well, the world functions around problems, not necessarily around solutions.” And I think that that shift only happened then. I wish, and if I ever start a business again, I will start with that mindset again. I didn't. But luckily, we are in this world that is changing by machine learning and all these companies are drowning in data. So we just landed on a solution that had a problem to it, and a problem that was very large.
Savodivker: Did you make a pivot when you found that, or what was the result? Because you had picked up some money before that, and then it seems kind of like you figured out maybe some changes that you needed to do. What went into that?
Torres: Yeah. So MindsDB didn't change that much, but the way that we see MindsDB and the way that we see even the vision of MindsDB changed. Even the idea of a vision of MindsDB long term all of a sudden changed because we could see, okay, well the world has this problem, and it can be different if we solve the problem at a large scale, even if the product as we have it right now can change.
To the point that I'm describing to you, we were focused only on the technical people that needed to apply machine learning. But the way that we see it now is, look, the technical amount of individuals that can use our product is a lot of them. But then there's all this other amount of individuals that work for companies that could be using this type of capability. And then even imagine a world where anyone that works in a company can use all the data to make informed decisions or better informed decisions, then all of a sudden it gives you a better North Star to pursue. So even the product as we see today can evolve to pursue that North Star.
Savodivker: So Jorge, you're in tech, you're obviously this phenomenally smart person that's working on machine learning. Can you talk to us about any biases that you've encountered as a Latino in tech, as you've fundraised or as you've gone through the growth of the company?
Torres: Even still today, people tend to reach out to Adam first.
Lantigua: And is that because he's white?
Torres: I've asked myself this question, and it is for trivial stuff. But yes, I do think that people are not used to a Latino CEO. They're not that familiar, but there are more. And so many of them now are getting super big in their own markets. But of course, it's not the norm. And I had this one experience, this is pre-COVID, and just let me put a side note. COVID has made it so different because you're behind the screen, and then behind the screen, then a lot of preconceptions about someone, the stereotypes that people have for what entrepreneurs should look like, are kind of removed. I think that Zoom just made the whole thing much more neutral.
Anyway, but long story short, I was going to the offices of this fund. I'm not going to say the name. And I show up, and I remember the lady in the reception, she's white. I don't think she's familiar with entrepreneurs or Latinos. And then she's like, okay, I'm here to talk to so and so. And she's like, "Okay, what are you going to deliver?"
It just struck me as wow. But then on the other hand, I was like, okay, this person hasn't seen someone like me clearly. I mean, she probably has seen a lot of people that look like me that go to deliver stuff. I think that it is an idea that people have in their heads, and it takes for an experience like this one to change their minds. I wanted to laugh, as well as to take it seriously, but all of a sudden it's just like, "No, I'm here to deliver a pitch." And then you could see the horror in her face. She was like, fuck, I screwed up.
Savodivker: And let me ask you something. So from that experience, did it change anything or did it have any impact on you as you continued to fundraise?
Torres: I think that it did play with my mind that I didn't fit the mold, but also at this time as well, it was when we were surrounded by so many entrepreneurs when we were doing the program at Stanford, and even at Berkeley, I got to meet so many other ones, that I felt I had enough support network to be like, okay, let me talk to people that may go through the same experience or not.
So I think that the right mindset is to assume that these people will invest in you because the problem is a fantastic problem and the solution that you think you have for this problem is the right one. And then the metrics that you have around it are solid. And you should focus on all of this.
And again, I understand I have an advantage as well because I have a business partner that is not Latin American in that sense, but I have met so many other entrepreneurs that are solo founders, female Latin American founders, and they're still killing it. And you can see them, and their journey is the same. And this story is you know what, screw that. The name of the game for an investor and the name of the game for people that go into business with you is to pursue great projects and great ideas and to have a win-win situation.
Savodivker: Well yeah, and I think it's helpful to have someone like you leading the way, to show that these problems are being fixed by Latinos, along with Latinas and just about everybody else.
Lantigua: Jorge, thank you so much for the tech lesson for me, but also for the really wonderful insight into how you innovate and how you persevere, and especially that wonderful breakdown of the process of being VC funded at launch. I think that that's going to be really helpful for a lot of people.
Torres: Thank you so much for the invitation. Take care.
Savodivker: Juleyka, what was your biggest takeaway from our conversation with Jorge? Because you were really taking it all in, huh?
Lantigua: I really did, though, because at heart I am a science geek. I really am. I love learning how things work. I love listening to people explain how they're going to build upon knowledge that we already have. And that's exactly what he's doing, right? Building upon the knowledge that is very rapidly being accumulated about machine learning and about AI.
And so what I will take away, though, is the fact that humans are still essential, and how we view the world individually and collectively shapes what the machines do. And so to me that says ultimately we have to maintain a high level of discernment and responsibility. The tech is really important, but the ethics of how we build that tech are just as important. And that would've been one of my other questions, but I mean, we could have talked to him for hours.
We might have to bring him back, and maybe we do a special and a couple of tech founders just to talk about the ethics of technology because I feel like as Latinos, we are very much rooted in family values, in a lot of communal thinking. In our culture, there are particular values that are built out of particular ethics. And so I would be really curious to understand how and if they're building those into the DNA of the tech and the AI that they're building. But I think what they're doing is so, so smart.
Savodivker: Juleyka, I agree. Jorge just does such a good job of explaining everything. I'm not a science geek, but I was even able to take it all away. And I really like how he talked about the business side of things. He really was able to explain to us how you can go from a presentation and getting capital just with that to the later stages and what he's doing now, which is the next challenge, that series A, that much more difficult level of getting capital by having a product market fit.
Lantigua: So actually, I want to add one thing that I didn't get to highlight during the conversation, but which is so important, I believe. He's really clear that there is a system, which is a VC funded system, that has its guidelines, it has its measurements, it has its KPIs. So that is one system for getting your business funded and launching your business, but it is not the only system.
And I think that's really important because a lot of people have mythologized the VC route, as it were. But there are millions of other ways, and there are countless other entrepreneurs that every day find a way to launch their businesses. So this is an amazing example of how to deploy that system with a high degree of excellence, but it is one system.
Savodivker: Juleyka, I couldn't agree anymore. Obviously this is one system, one type of way of growing your business. What I really loved about Jorge's interview is that he talked to us about how he really strategized. He learned the game, he learned how to play it, and is maximizing the results of that. And I think that's something that you can take into whatever strategy you move forward in your business. And so just could not be any more happy with some of the things that he talked about. I took notes, and hopefully our listeners did, as well.
Now, here's our Latino business moment of zen, inspired by our guests. Inhale and exhale. Close your eyes and repeat after me. I am significant. I am needed, worthy. I am noteworthy. I contribute my significance to the world by being who I am and doing as I do.
Among those surveyed for LBAN's research, 19% of Latino owned businesses report developing and selling a technology or software product, while only 14% of all white owned businesses surveyed could say the same thing. That's 19% to 14%. The US Census annual business survey looks at the share of businesses in technology production. It reports that among Latino owned businesses, 10.6% are producing technologies such as artificial intelligence, specialized software, robotics, and specialized equipment. This 10.6% is compared to 10.1% among white own businesses doing the same.
You are significant, needed, worthy. Inhale and exhale. Now, open your eyes and carry on with your day.
Lantigua: This episode was supported by our LBAN Alumni Founding Sponsors Windrose Vision, VND, Fig Factor Media, Allegiant Electric and Global 4PL.
Scaled: The Latino Business Story is produced by LWC Studios for LBAN. Virginia Lora is our producer. Kojin Tashiro and Elizabeth Nakano are our mixers. Kojin Tashiro did our sound design. Paulina Velasco is managing producer. To learn more about the work and research LBAN is doing and our Business Scaling Program at Stanford, please visit LBAN.us, that’s l-b-a-n [dot] us. Thanks for listening. I’m Juleyka Lantigua.
CITATION:
Lantigua, Juleyka and Elian Savodivker, hosts. “Investors Want a Big Vision, Good Metrics and a Problem You Love.”
Scaled: The Latino Business Story,
LWC Studios., January 16, 2023. LBAN.us/scaled.