A Conversation with Alexandra Johnson, CEO/Co-Founder of Rubber Ducky Labs, AI Recommender Systems Pioneer

Laura: Alexandra, you've worked in machine learning for a while now, but we're at this new moment of inflection with AI and machine learning in the present tech environment. What feels different to you about right now?

Alexandra: What I'm most excited about, when I was reflecting on this for the current AI boom, is that there's kind of three stages of project development for any sort of project that you're going to go under. There's prototyping, there's bringing something into production and there's optimizing something that's already mature. The current AI boom hits the prototyping stage really well. People are able to get these applications that use machine learning up and running way faster than they ever could. I'm thinking specifically about the 2017 deep learning boom when everyone was kind of just running around thinking, "How do we even get ML off the ground?" With the current AI boom, you can get your prototypes off the ground way faster with way less need to hire specialists. That means that a lot more people can be brought into the process and a lot more people can bring their vision to reality.

Laura: That's fantastic. What types of problems are you trying to solve today at Rubber Ducky Labs?

Alexandra: We're really focused on the problem of recommender systems, which is showing the user the right products at the right time. This is really driven by a personal love that I have for online shopping. I grew up in a rural area where there wasn’t a shopping mall anywhere nearby. You'd have to drive two hours to the nearest shopping mall. I fell in love with online shopping early and it was a personal passion of mine to help make online shopping better. I discovered recommender systems in college, and it was really apparent to me that machine learning and a lot of the new data mining techniques that were coming up were going to be a really impactful way to touch online shopping and to create really exciting shopping experiences. That's been proven out with different new technologies, like TikTok in recent years, that's completely upending the discovery process for how people come across clothing and trends and lots of other products online. With what we're doing at Rubber Ducky Labs, it's just quite simple. Like I said, helping people find the right product at the right time through the use of machine learning and AI-powered recommender systems.

Laura: I know that investors in the current market talk a lot about signal and noise. There are so many machine learning and AI companies out there. And you've had great traction with seed investment, with early customers and pilot users. What do you think differentiates RDL and you as an entrepreneur in this area?

Alexandra: Well, one is the hat. We have great hats. The company was named because I have a red hat with a rubber ducky on it. I've had this hat for 10 years, and deep in the middle of the pandemic was when I founded the company, and I was casting about for some sort of all business name. I was like, nope, the logo is going to be a rubber ducky. So there's that.

But then also we kept a focus on the recommender system space for a while. And there's kind of a few things about that that helps set us apart. One is that recommender systems are actually a slightly differently shaped problem from other problems in machine learning. So you'll have things like classification problems where you're like, is this answer going to be a true or false? Is this person going to repay this loan or not? You're going to have a regression problem where you're like, I'm Netflix. What star rating is someone going to assign to this movie? Or what loan amount should I approve someone for? You're going to have things like these text generation problems. And then you’re going to have things like recommender systems, which in itself, each one of these could be its own specialty. And so by specializing in recommender systems, that lets us build expertise up in the area specifically when it comes to solving those problems for our customer. And so that's something that that focus and expertise helps set us apart.

Laura: That's very cool. And bonus round, would you say that the proliferation of large language models and more and more AI-driven, open source becoming available is accelerating your growth or do you feel like you're really building on the fundamentals that you've been building on since before the sort of large language model revolution?

Alexandra: Yeah, you know that we're really focused on fundamentals. We're really focused on saying, what's the problem we're trying to solve for our customer? Which is increasing revenue, increasing engagement through the use of recommender systems. And then as with any engineering project, how can we use the right tool for the job? So I actually started fundraising like right before the boom. It felt like it was days before the AI boom. And then this AI wave came up during the middle of the YC batch. So, it was just very interesting to look and to try to predict, like, where do I think the trends are going to go? What do I think these technologies are going to be used for? And I've seen over the last 18 months so many really interesting use cases and so many opportunities where I'm like, “Oh, hey, I want to see, where does that go? Where does that prototype go? Like, how does that end up making it into production?” And so, again, excited to see where all these things go and excited to see what bits and pieces we can pick up when it's appropriate and when it helps us solve our problem for our customers.

Laura: That's fantastic. So much gratitude for taking the time today. And I'm going to have to get a rubber ducky hat. That is the only next step here! Thank you.