Jamshid A. Vayghan, Ph.D

Q: What are your current research interests? 

My current work in industry and research interests focus on how organizations can successfully  scale AI and data-driven innovation from experimentation to enterprise-wide impact. While there  has been a surge in AI and data science pilots, many of these efforts struggle to generate  sustainable value at scale. 

I’m particularly interested in identifying and addressing the barriers to scaling, including gaps  in architecture, governance, skills, security, operating models, and organizational readiness. My  focus is on: 

• Designing scalable data and AI platforms that balance innovation with reliability and  control 

• Defining industry use cases that tie AI and data investments directly to business value • Enabling cross-functional collaboration so that AI capabilities are embedded into the  workflows of business teams—not just developed in isolated labs 

• Building feedback loops between experimentation and production that ensure alignment  with evolving business needs 

• Training the next generation of data scientists, equipping them with both technical  depth and business context 

• Supporting cross-disciplinary collaboration among business, computer science, law,  and public policy communities to tackle complex data and AI challenges 

Ultimately, my goal is to help enterprises turn promising ideas into durable capabilities that  drive measurable outcomes. I view this not just as a technical challenge, but as a strategic one— requiring a blend of engineering discipline, systems thinking, and organizational change. 

Q: How do you define data science? 

To me, data science is not just a set of statistical or machine learning techniques—it’s a holistic  discipline that spans the entire data lifecycle, from creation and collection to transformation,  governance, and ultimately consumption. It’s about turning raw data into trustworthy, actionable  insights that can be used consistently and responsibly across a variety of AI systems. 

During my work leading the transformation of data platform in a large global enterprise, I came  to view data science as a strategic capability—one that enables organizations to build the  foundation for AI and digital innovation at scale. This includes not only advanced modeling and  analysis, but also the infrastructure, quality controls, security and privacy, metadata  management, and governance mechanisms that ensure data is fit for purpose. 

In essence, data science is what connects data to decisions. It’s what makes it possible to go  beyond dashboards and predictive models to deliver real, measurable value—whether through  automation, personalization, optimization, or entirely new business models.

Q: Can you share an interesting or surprising result you’ve found in your data? 

Absolutely. One of the most surprising—and humbling—insights came during my engagement in  a data transformation initiative in a large global enterprise. We had invested in building  centralized repositories for key enterprise data subject areas, aiming to eliminate redundancy and  reduce infrastructure costs. On paper, we had created “trusted sources” of truth. 

But when we analyzed how the data was actually being consumed across the business, a very  different picture emerged. 

We discovered that teams were indeed pulling data from these central repositories—but then  modifying, reshaping, and storing it locally to fit their own needs. In effect, we had centralized  the storage but not the trust or consistency of how the data was used. The net result? While we  achieved some infrastructure efficiency, we fell short on delivering real business value or  enabling enterprise-wide consistency in decision-making. 

That insight led us to rethink our approach entirely. We realized that solving for storage and  access wasn’t enough—we needed to shift to an outside-in perspective focused on governance,  usability, security, and end-to-end consumption patterns. Only by ensuring that data was not  just stored once, but also understood, governed, and used consistently across the enterprise,  could we truly unlock its value. 

This experience reshaped how I think about data transformation—not as a technology project,  but as a systemic change in how an organization builds trust, drives alignment, and creates  impact through data. 

Q: Are there any new tools or libraries you (or your students) have been using lately? 

Rather than highlighting a single tool or library, what I find most exciting lately is the way  enterprise data technologies are starting to come together in more integrated and  interoperable ways. 

Platforms from vendors like SAP, Databricks, Microsoft, Oracle, Google, and IBM are no longer  just standalone systems—they’re evolving to become part of a more unified data ecosystem,  making it easier for large organizations to build end-to-end pipelines that connect data  ingestion, governance, transformation, and AI/ML deployment. This is especially critical for  enterprises with complex, hybrid environments. 

In parallel, I’m seeing growing momentum behind open-source technologies—from Delta Lake  and Apache Iceberg to tools like dbt and MLflow—which are being adopted not only by startups  but also by large enterprises as part of their modernization strategies. These tools are helping democratize access to advanced capabilities while preserving flexibility and reducing vendor  lock-in. 

The real story, in my view, is the blending of proprietary and open ecosystems to support  scalability, governance, security, and AI-readiness. That convergence is shaping how we train  future data leaders—emphasizing not just technical tools, but architectural thinking and  platform strategy. 

Q: What excites you most about the future of data science in the next five years? 

What excites me most is how data science is evolving into a more integrated and applied  discipline—closely intertwined with AI and business transformation, deeply grounded in  real-world use cases, and increasingly interdisciplinary in nature. 

We’re already seeing this shift in action. For example, at the University of Minnesota, the  original data science initiative has now expanded to include AI as “Data Science and AI Hub”,  reflecting the growing recognition that AI and data science are two sides of the same coin.  This integration is accelerating the development of skills, assets, solutions with tangible  industry impact, from personalized healthcare and smart infrastructure to financial risk  modeling and public policy. 

At the same time, data science is no longer just a subfield of statistics or computer science. It's  becoming a collaborative discipline that brings together experts from business, law, public  policy, and engineering. This kind of cross-functional partnership is essential—especially as we  grapple with challenges like data privacy, algorithmic bias, and responsible AI. 

Looking ahead, I believe security will play an even more central role in data science. As data  becomes more distributed, valuable, and sensitive, ensuring that it's protected, trusted, and used  ethically will be as important as the models we build on top of it. 

To me, the future of data science is about building trustworthy, intelligent systems that are  not only innovative but also accountable and inclusive—and that’s both an exciting and  important challenge to work on.