Data science is a profoundly interdisciplinary field, which makes it difficult to define precisely. It’s a profession that entails more than just coding—the daily tasks and responsibilities of a data scientist can vary drastically as can the use of data, depending on the company.

By 2025, it’s estimated that 463 exabytes of data will be created each day globally, according to Raconteur. More data provides opportunities for improving business strategy—but the more data available, the more personnel are needed to manage, analyze, and create solutions from it.

That expanding need for data scientists is something Mohamed Abdelsadek knows quite well. He oversees data science efforts at Mastercard as the executive vice president of insights and analytics in the company’s data and services division—which, along with cyber and intelligence services, generated 35% of Mastercard’s net revenue in 2021. 

Mastercard’s data and services unit utilizes the data the credit card company collects to provide business tools and solutions to customers. These services were once directly tied to the company’s core business: its card products. However, the company began seeking ways to help customers effectively use all of the the information and technology Mastercard could provide.

“Over time, we’ve invested a lot in making sure our data can be used and consumed, so that we could provide additional value,” Abdelsadek says. “As we did more and more of these solutions, we started to evolve increasingly outside of the core business to be able to offer solutions more broadly.” 

The data and services team at Mastercard now provides solutions not only to banks and retailers, but also to small businesses and governments. Today, data science is widely applied at the company as data scientists work on product development, analysis, and customer support—ultimately turning data into actions.

The team is now supported by more than 2,000 data scientists, engineers, and consultants who come to the company with degrees in fields like data science, business analytics, information systems, mathematics, statistics, and engineering. Abdelsadek holds a master’s degree in computer science from Columbia University and an MBA degree from Wharton.

To find out more about what a day in the life of a data scientist looks like at Mastercard, Fortune spoke with Abdelsadek and other data scientists at the company.

It makes sense that a multi-faceted occupation entails a variety of responsibilities every day, but some tasks and schedule items are constants for Abdelsadek.

“If every day was the same, I’d be pretty bored,” Abdelsadek says. “My time is spent between customers, our partners, and our teams and employees.” 

Aspects of his routine remain the same though, and he finds it important to rely on some consistency to provide structure for the varied items on his to-do list. Every day, Abdelsadek wakes up between 5:00 and 5:30 a.m. and heads to the gym before work. Before he even turns on his computer, he writes down everything he wishes to accomplish for the day on a blank sheet of paper.

“I find that if I don’t do that, I just end up getting consumed with emails and everything that comes at me during the day and I end up not doing the things that I think are really important,” Abdelsadek says. 

Here are some of the items that are written on that day-determining piece of paper:

1. Meeting with customers: Abdelsadek typically has at least one meeting each day with a customer discussing requests and current product offerings.

2. Developing new product ideas: Abdelsadek communicates with several different Mastercard teams about new ways to utilize data and what product developments that are on the horizon. 

3. Mentoring others: Time is set aside for one phone call every day with a junior Mastercard employee to discuss ideas and career-related topics. 

4. Meeting with industry experts: As an example, Abdelsadek meets with Bricklin Dwyer, chief economist and head of the Mastercard Economics Institute, to look at trends that are impacting partner companies, as well as Mastercard’s own business. 

5. Determining key priorities for the business: A lot of Abdelsadek’s time is spent ensuring that the company continues to think about what third-party information—in addition to its own information—can help complement and make Mastercard’s services more powerful. One way he does this is by routinely working with JoAnn Stonier, Mastercard’s chief data officer, to improve the company’s data infrastructure and analytical environment.

At Mastercard, data scientists have to balance the now with the new—meaning, they need to ensure the effectiveness of their current tools while preparing new tools for problems and topics they have not yet tackled. 

“We have over 2,000 consultants, data engineers, and data scientists who get deployed to help customers on a variety of different questions,” Abdelsadek says. “For some of these products, we’re working with customers for years at a time on a variety of different problems.”

As Mastercard continues to expand community-driven partnerships, the company is focused on growing its data science team to keep up with additional projects. In February, the data and services team announced plans to add more than 500 college graduates and young professionals.

The day-to-day tasks reflect an interdisciplinary nature of data science, according to Fuyuan Xiao, a data scientist and a director of product management at Mastercard. While new solutions are always on the docket, Mastercard data scientists tend to rely on a consistent system to work through new areas of development.

“Data science’s nature is to use techniques drawn from many fields including statistics, mathematics, computer science, and information science to develop algorithms that can extract insights from data,” Xiao says. “When we apply data science to solve industrial problems, domain knowledge is also required.”

Xiao’s day-to-day workflow is a cycle of four steps. First, she defines business goals by considering the clients’ needs and consulting experts. Second, a statistical model is designed, determining the best type of algorithm to address the business problem. Third, Xiao and her team train and validate the model with massive amounts of data.

“Model training and validation will happen in a recursive manner—so it’s important to use the proper programming tools and systems,” Xiao says. 

Finally, the team applies what they’ve learned to solve the determined business problem and help their client meet their the goals. 

“Many data scientists in the field know it’s not enough to be very good at coding, or interpreting data, or creating automated data pipelines, or machine learning, or speaking,” says Joel Alcedo, a data scientist and vice president of applied economics at the Mastercard Economics Institute. “The most successful data scientists I have seen are equally good at all of it—kind-of like a Swiss army knife with various tools they can apply at any given time.”

Data scientists don’t merely handle technical issues and provide code for products; they frequently interface with internal and external stakeholders.

Any data science-related solution involves collaboration across several departments—including client services, product development, business development and account management, and more, says Vishal Arora, a data scientist and senior managing consultant for Mastercard Advisors’ client services team. 

So while the daily routine of a data scientist varies drastically, essentially every day includes a mix of technical tasks, analysis, and communication.

“Effective data scientists also know how and where there are gaps in their team’s workflow and work with the right stakeholders to improve the team’s efficiency,” Alcedo says. “They also know where big gaps in their own skills are that they can learn from and lean on others for support.”