The Berkeley School of Information historically admits only 30% to 35% of applicants each year who are either transitioning to a data science role or who are looking to hold leadership positions in the field, Alex Hughes, the head graduate adviser for the program, tells Fortune.

The average incoming student has a 3.5 undergraduate GPA and is in the 79th percentile for their GMAT score, although a GMAT or GRE score is not required to apply.

“There are other programs that are also very good, but they have perhaps a slightly different focus than we do,” Hughes says. “We’ve got a view that data science is more than just sitting and programming—that data science involves issues of identifying the core question that’s coming from business and being able to focus that into the type of question that we can bring an algorithm to solve.”

Not all candidates come from a programming or data-related background, but they all have to demonstrate an ability to work hard on a problem from inception to implementation, Hughes says. Successful candidates understand “that data science is more than simply programming or coding or stats or math—and instead, it’s a broader whole enterprise that has data as its core focus but has a lot of dimensions to it,” he adds. 

Hughes and current data science master’s student Srishti Mehra shared with Fortune some insight about the application process to UC Berkeley—and what sets the program apart.

Know what your goal is

The university means it when it says candidates don’t need to have prior programming or analytical experience before applying. Hughes says that UC Berkeley “intentionally keeps a mix” in the backgrounds of its admitted students profile. 

Learn more: Is a master’s degree in data science worth it?

About 5% to 10% of the class holds a Ph.D. in another discipline like computer science or economics. The remainder of the class is made up of students who hold an undergraduate degree in a cognate field like computer science, economics, political science, or psychology; professionals who are looking to make a career transition into data science; and those people who are looking to better manage data science teams.

While the cohort comes from different backgrounds, they have one thing in common: They know what their career goal is after the program. Applicants must express their intent and vision through a statement of purpose, which is part of the application.

“We read for the statement of purpose—that the applicant has some vision for what they want to do with their education when they’re through,” Hughes says. “We understand that that changes as people go through the program, but that they have some idea of what they want to do with this.”

Mehra knew exactly what she wanted coming into UC Berkeley’s full-time data science program. She earned her undergraduate degree in computer science and spent the first four years of her postgraduate career working as a software and data engineer at Goldman Sachs. After a while, she knew there was more that she wanted out of her career in data. 

“Being close enough to data, I realized there was a lot that I could do with it,” she tells Fortune. “I was a data engineer by title but [assumed] a lot of data analyst responsibility, as well, because of how close I was to the data and how much I knew we had and what could be done with it potentially.”

She then decided that she wanted a formal education in data science to pursue a career more focused on machine learning and artificial intelligence, which she addressed in her statement of purpose. Mehra started the master’s program in 2020 and will graduate this year—and still intends to pursue machine learning and artificial intelligence engineering roles. 

“Students should be focusing on their statement of purpose,” Hughes adds. “They should have a clear vision—or at least beginnings of a vision—for how they want to use data science in work that they do.”

Recognize your strengths and show you’re a problem solver

Mehra recognized that going into the program, one of her strengths was that she had an undergraduate degree in computer science and plenty of exposure to coding and working with data. She has friends in the program, though, who came from different academic backgrounds, including physics and engineering, which, she says, has been beneficial to her experience at UC Berkeley.

“The school has been good enough to capture different strengths, but all of those strengths can be applied to the program,” Mehra says, adding that working on projects with fellow students showed her that every person on her team had something to add that the others didn’t.

To be successful during these projects and other data science curricula, students have to be strong problem solvers. 

“One of the things that separates a successful data scientist from others is they’re willing to work the entire way through a problem—not just 80% of the way through the problem or not just 90% of the way through the problem—but to really think through every facet and every twist and turn through the entire problem,” Hughes says. 

Applicants can do this by discussing how they’ve been deeply engaged in a months-long problem and were able to solve the problem or bring change forward to the organization, he adds. This can be evidenced through letters of recommendation, the statement of purpose, and interview with the admissions committee.

Have the ‘resolve to work hard’

The UC Berkeley online master’s program in data science takes 20 months to complete, and most people who participate in the program are also working professionals. In your application, you need to demonstrate that you have the “resolve to study hard” and show you have the capacity and interest to put in the work, Hughes says. 

Mehra not only puts in the work in her classes but also serves as a teaching assistant (TA) for the program. She says the professors at UC Berkeley inspired her to take on more responsibility during her time in the program.

“I see a genuine excitement in them to teach and to help us learn, and I try to extend that now as a TA,” she says. As a TA, Mehra takes time to help prep materials for people who come to office hours to learn additional curriculum outside of the classroom and answer questions for projects and other assignments. 

“The more time you put in, the more you get out of it,” she says about the program.