Teaching Data Science in the Age of AI
With more than a decade at Google leading data science initiatives in measurement, experimentation, machine learning and marketing analytics, Gülşah Yilmaz Benk brings academic rigor and real-world expertise to our Introduction to Data Science class.
Alongside her industry work, she earned a Ph.D. in management information systems from Boğaziçi University, focusing her research on data science, machine learning and customer analytics. Throughout her career, Gülşah has tackled everything from customer lifetime value modeling to AI-powered solutions, fueled by what she describes as a love for curiosity, creativity and problem-solving.
Since joining our esteemed roster of instructors in 2025, Gülşah has found a natural fit among the diverse community of learners. She particularly values helping students connect academic concepts with the challenges they'll encounter in the workplace, bridging the gap between theory and practice through hands-on learning and real business examples.
How would you describe your students?
They are incredibly motivated.
Many are balancing careers, families and personal commitments while investing in their education. Others are changing careers, returning to school after years away or learning new skills while working full time. That takes courage, and it's something I deeply respect.
I like to tell students that data science is part detective work, part storytelling and part problem-solving. It's not just about building models. It's about asking the right questions and turning data into meaningful insights.
Tell me about the topics you discuss in your class?
We cover the full data science journey, from collecting and cleaning data to exploring, visualizing and analyzing data, all the way to building introductory machine learning models. Students also learn statistical thinking, how to communicate insights effectively and how data-driven decision-making works in practice.
I like to tell students that data science is part detective work, part storytelling and part problem-solving. It's not just about building models. It's about asking the right questions and turning data into meaningful insights.
Whenever possible, I connect concepts to real business problems I've encountered. I also share examples from real-world projects involving experimentation, predictive modeling and AI applications. My goal is to help students understand not only how these techniques work, but also when and why organizations use them to make decisions.
How do you support students during the class?
I create a classroom where questions are encouraged rather than avoided.
If one student is confused about a topic, there's usually a good chance several others are wondering the same thing. I want students to feel comfortable speaking up, asking questions and actively participating in discussions.
I also find that confidence grows quickly once students start working with data themselves. That's one of the reasons we incorporate hands-on labs throughout the course.
Many students may feel intimidated by data science. How do you make the material approachable for beginners or those who have dabbled a bit in data?
I remind students that nobody starts as a data scientist. Data science is a skill, not a talent.
When people see polished dashboards, AI models and advanced analyses, they often assume experts understood everything immediately. The reality is that everyone starts by learning one concept at a time.
I focus on building intuition first. Once students understand the "why," the technical details become much less intimidating.
I also find that confidence grows quickly once students start working with data themselves. That's one of the reasons we incorporate hands-on labs throughout the course.
Can you share a memorable student success story or breakthrough moment from your class?
I remember one student who was initially nervous about coding and unsure whether data science was the right path. By the end of the course, this student completed a final project on a cared-about topic and presented the findings with confidence. Seeing that kind of growth over just 10 weeks is incredibly rewarding.
Moments like these stay with me.
I've seen data science transform industries such as health care, finance, retail, manufacturing and education. Organizations increasingly rely on data to improve decision-making, optimize operations and better understand customer needs.
Data science is everywhere. How have you seen data science implemented within tech specifically and more broadly in other industries?
Data science touches almost every decision we make today, often in ways we don't even realize.
In technology, it powers recommendation systems, search ranking, experimentation platforms, forecasting, personalization and AI applications that millions of people use every day. At companies like Google, data science helps teams understand user behavior, measure impact and make better product decisions through experimentation and evidence-based insights.
Beyond tech, I've seen data science transform industries such as health care, finance, retail, manufacturing and education. Organizations increasingly rely on data to improve decision-making, optimize operations and better understand customer needs.
The fun part is that once you start thinking like a data scientist, you begin seeing data science everywhere.
Data science is driven by AI tools. What other skills do students need to hone in order to collaborate with an AI tool?
AI can help generate answers, write code and automate parts of the analytical workflow, but humans still need to know whether those answers make sense.
Students still need strong analytical reasoning, statistical thinking, problem-solving skills and domain expertise. Just as importantly, they need to know how to ask good questions and evaluate whether an answer is actually correct. AI can generate outputs, but it can't replace human judgment.
What’s something students can do after your course that they couldn’t do before?
By the end of the course, students can take a messy dataset, clean it, analyze it and build an introductory machine learning model. They learn how to communicate complex analyses and insights to people with different levels of technical expertise, including decision-makers, stakeholders and non-technical audiences. That's one of the most valuable skills a data scientist can develop.
What advice would you give to your student on how to best succeed in our current workforce?
Stay curious and adaptable.
The technologies will change. The tools will change. What matters is your ability to learn continuously, solve problems and adapt to new challenges. I also encourage students to invest in their communication skills. The ability to explain complex ideas clearly is often just as important as technical expertise.
Outside of work, where can we find you?
Ever since I was young, I've loved being near the sea and imagined living close to it. There's something about the sound of the waves that helps me slow down, clear my mind and think more clearly. It's where I often reflect, plan my week or simply recharge.
I also enjoy exploring different parts of the Bay Area, traveling and discovering new places.
What is the one item in your office that is most representative of your personality?
A custom LEGO figure on my desk that says "Dr. Gulsah." It reminds me of some of the important milestones in my academic and professional journey and how much persistence it took to reach them.
