Defining the AI Stack: A Layered Approach to Data Science and Artificial Intelligence Education

This article explores the seven interconnected layers that comprise the AI stack — and how they collectively benefit those looking to advance in the field. 

“Most people, when they think about AI education, they think about modeling. That’s a critical part of AI. However, it’s not complete,” explains Bruce Huang, director of information technology programs at HES. He believes students shouldn’t merely learn how to use AI; they need to understand how to build, design, and adapt the complete ecosystem. 

“We’re no longer asking AI to write emails for us. We are using it as a collaborator for our everyday workflow,” adds Huang. “To accelerate in the industry, you probably want to focus on how to ask better questions. To know when to trust AI, and when to question AI.” 

A Modern Education Framework for Modern Leaders 

Artificial intelligence has been embedded in Harvard Extension School’s data science master’s degree program for years. In the 2026–27 academic year, the title was updated to more prominently reflect the breadth and depth of the curriculum. 

“The future will not belong to those who simply know how to prompt a model. It will belong to those who understand the complete system,” says Huang. “The new name represents that spirit, and that’s also the spirit behind the HES-Defined AI Stack framework.”

Here’s a closer look at how the HES-Defined AI Stack framework informs the Data Science and Artificial Intelligence Master’s Degree Program curriculum. 

AI Stack Layer 1: The Foundation 

Building core skills is essential to leading in the AI field. This first layer builds a foundation in quantitative, statistical, and computational knowledge. With this training, Huang says students are prepared to develop machine learning and AI models.

AI Stack Layer 2: Machine-Learning and AI Models

At the machine learning and AI model layer, students start building linear regression models, solving classification problems. These skills allow students to create decision-making tools that help their organizations make better data-driven decisions. 

AI Stack Layer 3: Data Engineering and Systems

Huang says many AI-focused degree and certification programs stop at building models. But at HES, the learning goes deeper. “We believe that to lead AI, students must understand how the entire AI system works,” he says. So this layer teaches students how to design, develop, and architect the data pipeline, architecture, and systems. 

AI Stack Layer 4: Platforms and Infrastructure

Models and systems aren’t everything, says Huang. You can have the best, most accurate, most reliable model, but it won’t scale if there are performance problems. The platforms and infrastructure layer provides insight into how to power and run those models effectively and efficiently. 

AI Stack Layer 5: Application Domains

With Harvard Extension’s focus on real-world learning — and with students coming from industries of all kinds — layer five is especially important. Through courses and capstone projects, students apply what they’re learning to specific business problems in areas such as healthcare, finance, insurance, and technology. Huang explains that combining AI skills with domain expertise will help students become better advisors to their organizations or clients.

His colleague and HES instructor Stephen Elston adds that mid-career professionals looking to move into data science and AI roles can bring incredible value from their existing expertise: “Established professionals have the deep domain experience required to understand what problems are important and how to make analytic results actionable and, therefore, valuable.” 

AI Stack Layer 6: Ethics and Governance

The Data Science and Artificial Intelligence Master’s Degree Program isn’t just about becoming an AI expert; it’s about becoming a responsible AI leader. That’s why layer six exists. 

“AI is powerful, but at the same time it can be dangerous,” says Huang. “We want our students to have an understanding of their ethical responsibility when developing, launching, and using AI model systems.” 

The ethics and governance course, for example, covers a range of topics including fairness, interpretability, security, accountability, and the social implications of predictive and generative systems. 

AI Stack Layer 7: Integration and Application 

The final layer of the AI stack builds off the rest, and it’s what will drive maximum impact for students as they become leaders. “It’s all about the ability to integrate AI successfully into a company’s culture, into a company’s process, into the company’s workflow,” Huang says.

Putting it all Together: A Data Science Master’s Program Built to Shape AI Leaders

“Our master’s degree program is called data science and artificial intelligence for a reason,” says Huang. “There are many other institutions offering master’s degrees in machine learning and artificial intelligence, in data science, in business analytics.” 

The HES program is far more comprehensive. Together, the stack is a powerhouse of AI knowledge. And, separately, each layer is designed around skills that can be immediately put into practice. And it’s all rooted in an integrated framework that provides direction for future data science and AI professionals. 

“[The HES-Defined AI Stack] is a roadmap for students going from the beginning — understanding how to do statistical analysis and how to write Python programs — to building the models and understanding how to make them better, and then what infrastructure it needs to run on, and then how to apply all those in a specific industry,” says Huang. 

AI Stack in the Industry: The Tools and Technology Driving Organizations Forward

Huang notes that the term “AI stack” is used more broadly in the industry. In a general context, it refers to how companies position their products and services within the AI ecosystem. While technologies, components, and functions will vary by company, a typical AI stack includes the following layers: 

• Infrastructure 
• Data
• Model development
• Model deployment
• Application 
• Governance 

The HES-Defined AI Stack framework that fuels the data science and artificial intelligence master’s degree expands upon this concept. It explore the elements of the literal AI stack students will encounter in the workplace today. And it focuses on the human aspect of leading in the AI era. And that, says Huang, is what sets HES apart. 

“We did not frame our degree program in the context of the current AI tools available in the marketplace. We did not frame our AI curriculum based on the popular platforms currently available in the marketplace,” he says. Instead, the program prepares students to continue developing solutions as technology evolves. “Adaptability is what makes our program stand out.”

Gain an Edge with the AI Stack Framework: Consider the Data Science and Artificial Intelligence Master’s at HES

As job descriptions evolve and new career paths emerge, it’s never been a better time to pursue data science and AI roles — and the upskilling opportunities to qualify for them. Huang reiterates that today’s employers aren’t just looking for how well someone can use a particular platform or tool: “The most valuable thing we can give to our graduates is the ability to adapt and adjust and innovate.”

You can develop that ideal blend of technical proficiency and leadership skills in the Data Science and Artificial Intelligence Master’s Program at HES. We invite you to learn more about the customizable curriculum, expert instructors, and admissions requirements.