AI is Changing Data Science. Here’s How Northwestern’s MSDS Program is Changing With It

As generative AI tools become increasingly common, many professionals are asking how AI is changing the field of data science—and what skills data scientists need to stay competitive.

In some ways, AI has radically transformed the industry in just a few short years; what used to be primarily a retrieval-based industry has now shifted toward a generative model. Other examples of how AI is changing data science include:

• AI-assisted code generation and bug fixes, allowing developers to move faster
• Synthetic data, which is enabling new advances in healthcare and finance
• Multimodal AI pipelines, increasingly used in large language model development

But does the rise of AI mean that traditional data science skills are obsolete? To find the answer to this question and more that current and future data scientists are grappling with, we spoke to five faculty members from Northwestern University’s master’s in data science (MSDS) program.

Here’s what they had to say.

How AI Is Becoming a Data Science Skill

Artificial intelligence is becoming an increasingly common part of the data science workflow. Tools powered by generative AI can help with coding, automate repetitive tasks, and speed up aspects of model development. At the same time, the core foundations of data science—statistics, data engineering, modeling, and interpretation—remain essential for building reliable systems and translating data into meaningful business insights.

Atef Bader, an SPS faculty member for 25 years, calls 2020 the “big bang” of generative AI. Despite the changes that have proliferated since then, the fundamentals of data science have remained unchanged. Likewise, Syamala Srinivasan, who has over thirty-five years of industrial and academic experience in data science, notes that AI is only one small part of data science, and that the foundations, such as the analysis of structured data, are the same as ever.

Because of this, early core classes in the MSDS program rely on manual coding practices to give students the opportunity to build their skillset from the ground up.

That said, faculty in the MSDS program are reengineering other courses’ breadth and content to help students incorporate AI into their toolkit—a necessity for all modern data scientists. This goes beyond classes that make up the AI specialization and extends to courses in all five specializations.

Recent examples of how AI is incorporated across the MSDS curriculum include:

• MSDS 440: Conversational AI Assistants: A recently added course on designing and building conversational AI systems, including large language model applications and natural language interfaces.
• MSDS 442: AI Agent Design and Development: A recently added course on developing autonomous AI agents that can reason, plan, and execute tasks across data and software environments.
• MSDS 476: Business Process Analytics: An existing course not centered around AI, but in which students use AI to brainstorm, generate synthetic data, and resolve technical and connectivity issues. As part of the course content, students also learn to evaluate and refine AI outputs.

Rather than being a tool that replaces modern data science skills, Bader calls AI a “productivity multiplier” to help students speed through automatable tasks and instead focus on higher-level problem solving.

How does a master’s in data science prepare you for AI-driven work?

The MSDS program prepares students for the world beyond the classroom through hands-on learning, which equips them for the ever-evolving roles of data scientists and opens new doors even before they reach graduation.

“So many people have textbook knowledge but lack the practical knowledge and experience,” Srivinasan recalls from her 22 years at Caterpillar, Inc. “That’s not the case with MSDS students.” Using industry-aligned case studies from well-known companies such as Nike, Uber, Peloton, and Allstate, throughout the program, students design end-to-end data science solutions that combine:

• Predictive modeling
• Multimodal data processing
• AI-assisted analytics
• Business process automation

Examples of Hands-On Learning in the MSDS Program

Many MSDS courses are based around single projects that students work on the entire quarter:

MSDS 459: Knowledge Engineering

In program director Tom Miller’s course MSDS 459: Knowledge Engineering, students take on the role of an analyst from a company of their choosing within 12 sectors. They then spend the duration of the class building a machine learning model using open-source data to generate a report on their sector and forecast the trajectory of their chosen company’s stock.

MSDS 453: AI and Natural Language Processing

This course includes three real-world case studies each quarter. By the end of the class, students gain the applied experience needed to work as NLP engineers.

MSDS 485: Data Governance, Ethics, and Law

In this MSDS core course, students build in-demand management skills that can often be applied immediately. “Whenever students take this course, by the end of the term, at least one or two will have been brought on by their organization to do data governance at work,” says Candice Bradley, who specializes in this topic herself.

How MSDS Faculty Keep Up With AI and Data Science Trends

A Commitment to Continuous Learning

Northwestern’s faculty members share a dedication to continued learning, which is a key factor in keeping the MSDS program up to date with the latest changes in the field.

The MSDS faculty keep on top of such a rapidly evolving environment in data science by:

• Following AI and data science news
• Participating in professional online communities
• Attending industry panels and conferences
• Developing new skills, including prompt engineering and AI agents

Some faculty members even formalize their pursuit of knowledge by becoming students themselves, like Abid Ali, who has just started a law degree in addition to his doctorate and three existing master’s degrees.

Learning Through Teaching

In addition to asynchronous course content, all MSDS courses feature at least three sync sessions during which faculty members share the previous weeks’ developments and help students integrate these findings into their work. 

Between sessions, the professors guide students in where to find knowledge and how to interpret it.

And of course, “when you teach, you learn,” Bradley says. “Every single student brings their own perspective and knowledge.” With so many MSDS students already being data science practitioners, they bring their own wealth of practical knowledge to the table; past graduates have even commented that what they learned from their peers was just as valuable as what they learned from their professors.

What Skills Help Data Scientists Stay Competitive?

Potent as AI can be, there are still many areas of data science that require human expertise. Employers continue to seek data scientists capable of delivering actionable outcomes from their work, rather than the general analysis that AI provides. As Ali explains, “It’s augmented intelligence, not artificial.” Like any other tool, AI is only as powerful as the person using it.

Faculty emphasize that long-term success in data science requires a combination of technical expertise and professional skills.

Technical Skills That Remain Essential in Data Science

Students in the MSDS program develop core data science capabilities such as:

• Machine learning foundations
• AI applications and model development
• Model evaluation and validation
• Data analysis and interpretation

Professional Skills That Drive Impact

Equally important are the professional skills that allow data scientists to translate insights into real-world results. These include:

• Communication and storytelling with data
• Collaboration and teamwork
• Confidence presenting complex findings
• Understanding business context and decision-making

Preparing Students for Real-World Data Science Roles

“The design of the program lends itself to student success,” according to Srinivasan. To foster networking and collaboration skills, for example, classes often feature group work for applied exercises. And as the final measure of their growth in the program, students deliver their capstone projects orally to a panel of faculty, simulating a workplace presentation.

By the time students graduate from the MSDS, they will possess not only the technical skills, forged in practice, for a successful career in data science, but also the business skills required to effectively implement their findings.

The program’s external advisory board once described the MSDS as “worth its weight in gold” in terms of graduates’ preparedness for their post-program roles. “Employers know from day one, they can give our students projects and they’ll take off running,” Srinivasan remembers board members telling her.

How MSDS Prepares Students for the Future of Data Science

Like digital transformation before it, GenAI-driven transformation is redefining data science through automation and intelligent pipelines, and every indication points to its continued evolution in the future. Companies like Meta and Google, for example, are already working on AI world models, which will be able to generate three-dimensional virtual environments with their own physics systems.

“None of this is news to us,” Miller says; since the program’s inception, it has undergone constant revision to keep ahead of changes in the industry.

It may not be possible to predict what comes next, but faculty are confident that MSDS graduates will rise to meet new challenges thanks to the flexible skills they gain in the program, including critical thinking, ethical judgement, and a continuous growth mindset. 

When asked about the most valuable skill he passes on to his students, Miller cited skepticism: “Always question what you see, what people tell you. Identify the source of your information and ensure that source is real and credible. Not just in data science, but in the public sphere.”

Ali is proud to impart thoughtfulness and problem-solving skills to his classes, and Bradley makes sure students leave her courses having developed a deeper sense of fairness and ethics.

From the deeply technical to the more human side of data science, the MSDS program encompasses all the tools graduates need to stay relevant in this dynamic field and lead their organizations to success—not only today, but for years to come.