Role Models in AI: Carolina Barcenas
Carolina Barcenas is the Senior Vice President of Research at Visa, where she uses machine learning and statistical modeling to develop the payment technology we use every day. Spurred into modeling by her desire to understand and depict the world’s natural phenomena, Carolina is vocal about the importance of channeling your energy toward the problems you’re passionate about.
Carolina is active in developing programs that attract and retain underrepresented populations to tech fields, particularly women and non-traditional students. As the former co-lead of Visa Women in Tech, she created a community based on shared experiences, where women learn the leadership and networking skills necessary to thrive in the tech field. She also used her influence to partner Visa with a local nonprofit, creating a pipeline for students pursuing their associate’s degrees to intern at Visa and other corporations.
Read on to learn more about Carolina’s work and career in fintech, as well as what she thinks we should be aware of as AI technology develops.
As told to Nicole Halmi of AI4ALL by Carolina Barcenas; edited by Camryn Burkins.
NH: What do you do in your current role as SVP of Visa Research?
CB: I have a team of researchers who do foundational work, data scientists who do more of the applied science work, and AI research engineers. We mainly focus on four areas: artificial intelligence, particularly around deep learning; blockchain; security, in particular, quantum security and encryption; and finally an area that we call the “future of commerce,” which includes things like new payment experiences leveraging Internet of Things. As a payment company, we are constantly looking at how to innovate and how to be prepared for the future. We do foundational research, as well as product innovation. Some of these products eventually become part of products that we sell or products that we use internally.
What are some of the unique challenges that fintech companies like Visa face when implementing AI and machine learning techniques?
The one that always comes to mind is the implementation of models, mainly because of the scale at which we have to handle the input. The Visa payment system can handle more than 60,000 transactions per second. That means we have a lot of transactions coming through and very little time to execute. We work on optimizing the models so that they’re as accurate as possible while completing transactions as quickly as possible to ensure the best and most seamless experiences for everyone using Visa. These models can have millions of hyperparameters and we only have milliseconds for scoring.
You were Austin’s co-leader of Visa Women in Technology and you are the organizing force behind a community college intern program that focuses on non-traditional candidates. Can you tell us more about your work to meaningfully include and celebrate people who have traditionally been underrepresented in technology fields?
The main motivation for my work is both attract and retain women in the field of technology. We believe that organizations such as Visa Women in Technology allow a community to be formed where we share common experiences.
I also feel that it can be difficult for underrepresented and non-traditional candidates to enter technology companies. I had the opportunity to work with a nonprofit that enables students getting their associate’s degrees to get internships with companies like Visa. The students work at Visa while getting their degrees, and by the time they graduate they can decide whether they want to pursue a career in Visa, if or they want to go somewhere else. And even if they go somewhere else, we believe that the internship experience really opens up opportunities for them, because now they have very specific technology experience at Visa and similar companies. Some of the students have stayed with us upon graduation, and they’re brilliant. They’re doing some amazing work.
You have a PhD in Statistics and a BS in Physics Engineering. How did you decide to start out in Physics Engineering? What sparked your transition to Statistics? And how did you come to focus on fintech in your career?
When I was young, I was very interested in understanding how the world worked, and I had this idea that studying physics would really allow me to learn how it worked. As I was getting my physics degree, I realized that what I was actually passionate about was modeling, which is the ability to take some phenomenon and represent it mathematically. I’m originally from Mexico, and when I completed my undergrad I received a Fulbright scholarship to get a PhD in America.
When I was deciding what to study, I settled on statistics because I wanted to do modeling. I wanted to be able to represent the world.
When I finished my PhD I looked into what areas I could go into and decided fintech was the place I wanted to be. I’ve been in fintech my entire career. If you’re in this area of modeling, what you’re modeling is people’s money and livelihoods. The accuracy of our models has a profound impact on people’s day-to-day lives. I find that incredibly motivating.
Do you have any role models?
It’s not a single person, but I really look up to a certain type of person — the type of person that is willing to speak up to protect and help others. I think about countries or governments where they don’t have the same freedoms and liberties that we enjoy, and how some people have actually made a point to speak up and hopefully make a change. I really look up to people like that.
What has been the proudest or most exciting moment in your work so far?
I was working on a project where there was little belief that it was addressing a true problem. We saw it as a problem, but very few people thought that it was an issue. We decided to pursue it using some combination of algorithms and techniques that were quite new. It turned out that once we continued building a solution, the prevalence of the issue became more obvious.
Many times we stop short of pursuing something that we believe, because everybody’s telling us there’s nothing there, or you shouldn’t spend your time or your energy. If you deeply believe in a direction, stick to it, because sometimes you’ll be surprised, and good things may come out of it.
I think it’s less about the technology and all the work that goes into solving a problem, and more about the learning and the importance of persevering.
What advice do you have for young people who are interested in AI who might just be starting their career journeys?
Make sure that you find the right problem to solve, a problem that you are really passionate about, a problem that you are really curious about. I believe that if you’re not interested and fully committed, you might not have the energy to persevere. So make sure that you are working on something that you are very passionate about.
Another thing I would say is that you should always challenge the status quo.
As a female, there’s a very different perspective that you bring, and because of your perspective, you will see new opportunities that have been missed in the past.
Also, it’s very important to network. I wish somebody had told me that earlier in my career. Find good mentors, because good mentors are role models with whom you can personally identify, and give you support in those critical moments when you don’t know what to do.
About Carolina
Carolina Barcenas is Senior Vice President of Research at Visa. She leads an organization that is responsible for innovation in Artificial Intelligence, Blockchain and Security. Her organization focuses on both fundamental as well as applied research and works with the Product and Business teams on the creation of early products that incorporate the research outcome.
Carolina has served as Austin’s co-leader of Visa Women in Technology as well as the organizing force behind the community college intern program that focuses on non-traditional candidates.
She has worked both in industry as well as academia and has over 20 years of experience designing predictive analytical solutions in fintech. Prior to joining Visa, she spent 7 years at PayPal where she was responsible for managing the risk of small and medium e-commerce sellers.
Her expertise is in deep data mining techniques. She holds a PhD in Applied Statistics from the Georgia Institute of Technology as a Fulbright Scholar.
