When Roberto Valdez (’15SPS, Actuarial Science) is asked to describe his career, he refers to his life as “The time before my first master’s from Columbia, and the time after.”
In 2011, Valdez graduated with his bachelor’s in mathematics and was promptly hired by a former professor to work full time for an actuarial consulting firm in Quito, Ecuador, where he had lived his entire life. His future seemed set, and he was even on a path to making partner one day.
But he quickly discovered that something was lacking.
“I never felt that I had a proper actuarial science background,” Valdez explained. “I was a math guy who knew finance. I soon realized there is much more to statistics, and much more to actuarial science. You need to understand insurance and how it works, you need to understand the laws of the country you are working in, you need to know a lot about contracts.”
He applied for and was accepted to the M.S. in Actuarial Science program at Columbia University School of Professional Studies (SPS) through the Fulbright Program, which requires its scholars to return to their home country and work there for three years. He figured he would get his master’s, return to Ecuador, and get back on track to making partner and building a life in his home country.
While studying at Columbia, he took advantage of the flexibility to take electives outside the program and SPS, and on a whim he took a class called Statistical Machine Learning in the Engineering School. It was 2014, and AI was still only a curiosity for the average person. \
“I was hearing this buzzword, ‘artificial intelligence,’ that everyone was talking about at Columbia,” he recalled. “I was told it is actually statistics combined with computer science. I already knew statistics, and I knew some programming, so I thought, ‘Maybe I can complement those skills by taking this class.’”
That class, and his education at Columbia, would change the trajectory of his life.
“I realized, ‘Wow, this is the future,’” Valdez said about the machine learning class. “I got to see the beginning of all the artificial intelligence we are seeing right now. How you can apply data to train intelligent statistical models to solve problems — specifically day-to-day tasks.”
Valdez graduated with his M.S. in Actuarial Science, and he returned to Ecuador and worked for his former professor’s company for the three years stipulated by Fulbright. He then returned to Columbia for another master’s degree, this time in Computer Science. And now, a little over a decade later, Valdez is working as a machine learning engineer for Google, fully ensconced in the world of AI.
While many machine learning engineers enter the field strictly through software engineering, Valdez found that his SPS actuarial training gave him a unique edge. Understanding probability, risk distribution, and complex statistical models at SPS meant he wasn't just writing algorithms at Google — he deeply understood the mathematical foundations driving them.
“When people ask me about the difference between the two fields, I tell them they are not that far away from each other,” he explained. “The models that you use for solving actuarial problems, for example calculating how many people will die in a year, are similar to the models you use for solving these machine learning problems. You use your training, but you pivot a bit in terms of what problems you are dealing with. The fundamentals are the same.”
Though he no longer works in the traditional insurance field, Valdez credits his SPS degree with opening the door to his technical career. His time in the Actuarial Science program provided the statistical foundation for his data work, while the program's flexibility allowed him to discover his passion for AI, ultimately setting the stage for his machine learning work at Google today.
How would you say your experience in Columbia’s Actuarial Science program has helped prepare you for your role at Google and your career in the AI field?
Number one, my education in the Actuarial Science program taught me how to dig into the fundamentals, how to read deeply technical books, and how to understand theory. The program teaches you to take what you learn and condense it into something specific, like an equation or a theorem, and then extrapolate from there to solve an exam or work on a project. At Google, I am asked to do the same thing: look at very technical things, like code or a paper, condense them into ideas, and then extrapolate from those ideas to solve an issue on a project we are working on.
Columbia’s Actuarial Science program also forced me to sit down with people and explain what I know and to sell myself. When I first arrived at Columbia from Ecuador, I was the type of person who would undersell my achievements. Through talking to my classmates, or getting feedback from my professors and the staff who ran mock interviews for us, I learned how to sell myself and what I can do, but not in an arrogant way. It became a muscle for me.
Finally, the program taught me to dream big, and that is one of the things at Google they value a lot. Where I grew up, the professional culture often encourages taking a safer, more traditional route. Once you reach a respected milestone — like becoming a partner at a firm — the expectation is often to embrace that security, settle in, and enjoy the life you’ve built, rather than continuously chasing the next big risk. But Columbia’s Actuarial Science program told me I can interview for and do an internship at AIG, which is a world-famous insurance company. It told me I can sit beside these renowned actuaries and ask them questions and even disagree with them sometimes. Now I use that at Google, I propose new things, I dream big, and I suggest new ways of doing things; suddenly you start tearing down these invisible borders instead of staying in your lane.
Do you have any advice for current or prospective students of the Actuarial Science program?
Take advantage of Columbia’s environment and all it has to offer. In my case, for example, I took a machine learning class, and that eventually led me to the career I have now. Being at Columbia is such a specific experience, the knowledge is in the environment, your professors, your classmates. And you have such a mixture of people in your classes. I remember there were students from the Journalism School taking computer science classes with me, people from the mathematics department taking the statistics class with me in the Applied Analytics program. This is such a plus because you can listen to what they are doing, understand, ask them about what problems they are dealing with, and then suddenly you start getting a whole picture on the vibe of their industry. That is a huge edge.
As someone who is working in AI are there any big changes or trends in your industry that you think students should be aware of?
Sometimes people feel like they are competing with AI. I am working with AI at my side, so if tomorrow AI became twice as smart, my job would become much easier and I would actually benefit. There is going to be a fundamental shift in what knowledge and tools are important to have. For example, until recently, if you knew how to code in Python, it was a plus. Now AI writes Python code directly, so pure coding skills are no longer a plus.
Instead, the important skill set is to be able to code in Python and also to know how to talk to these AI models and iteratively build solutions based on a conversation with the tool. You still need to know the fundamentals so you can correctly prompt a model, decide the best path to follow, and know when the model makes a mistake. But you also want to be able to say, “I know how to interact with these programs, I understand them and I know how to use these agents to build solutions and to verify my work.” You want to demonstrate that you can work alongside them. “You aren’t just hiring me, you are hiring this entire network of tools and abilities I have that will enable me to be 10 times more productive.” That ability is going to be in very high demand.
Do you have any favorite memories from your time at Columbia and the Actuarial Science program?
I have one: When I was taking my final during my last class in my last semester, I remember the professor saying, “Hey guys, this is probably your last exam, you guys are graduating, so congratulations!” And I distinctly remember thinking to myself, “I hope it’s not my last exam. I need to come back here.” I loved my time at Columbia so much that I sincerely hoped that wasn’t my last exam. When I told my wife, who was my girlfriend at the time, she said, “You are so weird!” and that I was the only one in the room thinking that.
About the Program
The Master of Science in Actuarial Science program at Columbia University is internationally renowned for its breakthrough curriculum and esteemed faculty. The program equips students with the tools, skills, and knowledge to excel in today’s rapidly evolving actuarial and related workplaces, with a course of study designed to anticipate and exceed industry needs. Students are prepared to assume leadership positions and meet ever-expanding opportunities. Columbia’s location in New York City, the financial and actuarial capital of the world, allows students access to the world’s foremost practitioners and leaders.
Learn more about the program here.
