Jair Verçosa

Jair Verçosa, SEP ’25, is a serial entrepreneur. “I really like the ability to shift things quickly; I’m constantly pushing forward,” he says. Over the years, his desire for change has made him an expert at identifying trends and bringing technology to what he calls “unsexy industries.” He identifies a need, teaches himself the necessary skills, then builds solutions — and companies — to match.

Whether he’s meeting with future clients, assessing current markets, or taking his daughter to school, Verçosa’s always thinking about how to best solve problems using tech. His journey as a self-taught technologist started in Brazil, where he learned to code by embedding with the software team at his father’s logistics company. He then created an AI-powered social network to help match students with jobs, a full decade before AI models like Claude and ChatGPT became popular. He joined Carta, the financial software platform, in 2015; while helping grow the company to a valuation of $3 billion, he also started Flieber, a company that uses AI to help e-commerce companies optimize their inventory.

Wanting to increase his global reach as an entrepreneur, Verçosa enrolled in the Stanford Executive Program (SEP) at Stanford Graduate School of Business in 2025. On the first day of the program, he met his latest co-founder; they bonded while problem-solving together. Now they’re building their company — currently in stealth mode — at StartX. Jair credits SEP with teaching him to think bigger and more strategically, and connecting him to a larger community. As someone who always likes to be in motion, he says it was just the push he needed before entering a new sector: the maritime industry.

You’ve been an engineer. You’ve co-founded some startups. You’ve been an investor. How do you think about the different chapters of your career?

Early in my career, I acquired a lot of experience. My father was a logistics manager in Brazil. He took me to the warehouse, introduced me to the tech team, and said, “Try to learn something.” I learned to code, thinking about real business problems versus trying to build an app for fun. My career in Brazil was about enterprise resource planning, logistics software, and some startups.

The mindset in the U.S. is very different: You have more access to capital and more willingness to take risks, and that completely shifts the way you think. After spending most of my career in business-to-business companies, I started investing in Latin American startups as a way to enter the business-to-consumer space.

You’ve co-founded several startups. What keeps you in the startup game?

At a startup, I get to own the direction that I want to go. If I make mistakes, that’s on me. But also, if I hit a home run, that’s on me as well.

When I joined Carta in 2015 to build the Brazilian office, Carta was 30 people and valued at $77 million. I started from the ground up, got the Brazil office to a little bit over 100 people. Then I became the head of engineering for Carta and got transferred to the U.S. in 2017. I grew the engineering team here and helped the company to get to a $3.4 billion valuation in 2020. I left because we were 1,600 people and the mobility wasn’t there anymore. I needed to go back to then move forward again.

How and when did you get into AI?

I started working with machine learning in 2012 when I co-founded a company called Zona Universitária, which started like a Facebook of Brazil. Then we pivoted to recruiting. Our thesis was: It’s better for businesses to hire students based on their soft skills rather than the hard skills on their resume. I started building algorithms to match the skills the companies were looking for and skills students had.

In 2018, at Carta, I experimented with machine learning models for document parsing. A year later, I co-founded Flieber. Our goal was to identify how much inventory an online retailer needed. I had to learn a lot of things about time-series forecasting, so I got deeper into machine learning.

I didn’t have any experience with algorithms, but having a startup was the perfect excuse to learn. My first goal was: Is there anywhere else in the world where people are building algorithms from time series? I found some open-source packages and started experimenting. I had to learn how to shape the data, so I started going to meetups and asking questions. I also tried to hire people who were more experienced than I was.

Entrepreneurs are always problem-solving. What was the most unusual or interesting breakthrough you’ve had?

I realized that I needed to understand the data from a human standpoint to create strategies for how my AI model interacts with it. If you put any data in, an LLM will pick whatever it finds interesting. After Flieber, when I was using AI to personalize marketing messaging at Cordial, we developed this concept called intentional bias. The LLMs behaved almost like my daughter when we were taking her to school: She would stop all the time to pay attention to rocks or look at a flower.

I wondered: Can I use that in my favor? I said to my daughter, “Look, there’s a flower around the corner that you really like. Let’s just run there.” She ignored everything else and went straight to the flower. At the time, she was 4 years old.

Historically, bias is something you try to avoid. But if you have an intentional bias, you can push a model toward a certain direction. I now intentionally architect my data, thinking: What seeds I can plant that will make the model sound really smart toward the goals my users are trying to achieve?

Why enroll in the Stanford Executive Program?

I wanted to do something global and a little more challenging.

I had a friend who did the Stanford Executive Program in 2024. I thought: This is a program where I can meet people from different industries. As an example, in my cohort, there was a CEO of a fashion brand from Nigeria, and he was expanding to London and Brazil. There was someone who was the CEO of maritime companies and someone who sat on the board of four public companies. It was exactly what I was looking for.

On day one, I met my co-founder for my current company.

Amazing. How did that partnership come together?

We were assigned to study groups, and he was in my group. Immediately, it was a really good match. We would walk around campus, talking about his industry and potential solutions I had in mind. We were always together: breakfast, lunch, dinner, always talking about ideas. Several times, we were in a conference room after hours, using the whiteboard and thinking through problems.

Was there a particular class or faculty member that had a transformative impact on you?

One of the professors who impacted me the most was Jesper Sørensen. He would ask: For something to happen, what must be true? You have to think backward and figure out: What are the steps to get there? This changed me completely. Even now, every time my co-founder and I discuss a strategy, we always stop and say: Okay, what needs to be true? We go to the whiteboard and say: This needs to be true. This needs to be true. This is how we can get there.

What’s on the horizon for you? You have a new company in the works?

Yes. My co-founder and I are working to bring AI to the maritime industry, where a lot of things are still done manually, to help make processes more efficient and effective. The company is still in stealth. But we are building it at StartX, a startup accelerator that is part of the Stanford Community.