UC Berkeley’s Data Science Undergraduate Studies (DSUS) program, home to the nation’s top-ranked undergraduate data science major, continues to evolve its curriculum with faculty and student instructors. In the Data 101: Data Engineering course, collaboration recently led to a new instructional tool developed by graduate student instructor Rebecca Dang, working alongside teaching assistants Jonathan Ferrari (B.A. ‘25) and Christy Quang (B.A. ‘25), to improve how students learn core data engineering skills at scale.
The tool, which streamlines how students write and test database queries, replaces an outdated workflow and strengthens the course’s ability to teach complex concepts to more than 400 students each semester. Dang, a fifth-year master’s student in Electrical Engineering and Computer Sciences, presented the system at JupyterCon 2025, reflecting growing national interest in Berkeley’s instructional technology.
Lisa Yan, an assistant teaching professor in the Department of Electrical Engineering and Computer Sciences who co-teaches Data 101, said that since data engineering is a relatively new and fast-evolving field, identifying the most effective way to introduce its concepts to undergraduates requires ongoing experimentation.
“There are so many ways to teach data engineering and students may not have seen the tooling before,” Yan said. “Working with our TAs allowed us to design a system that teaches core skills in a way that’s accessible and manageable at a large scale.”
Although highly technical, the system was developed to solve a pedagogical problem. Moving query-writing into a separate file, rather than directly in a Jupyter Notebook, gives students clearer scaffolding, reduces common errors and makes it easier to teach concepts like query optimization.
“At the end of the day, it exists for a pedagogical reason,” Yan said. “We want students focused on learning, not on battling infrastructure.”
This student-centered approach is characteristic of DSUS, which gives student TAs significant responsibility in shaping course infrastructure, building tools, identifying pain points and proposing solutions used by thousands of students each year. Encouraging student innovation under faculty mentorship is a strategic element of the program, designed to cultivate emerging talent while strengthening the curriculum.
For Yan, developing future innovators and leaders is precisely what defines DSUS.
“The value of our program has always been teaching and developing strong undergraduate data scientists,” she said. “Our students help us maintain that mission. They push us to improve and they help us build the systems that make that improvement possible.”
The model also allows DSUS to keep pace with a field that changes quickly. Dang noted that continual iteration is a natural part of technical education.
“New versions of tools come out, error messages change, workflows evolve,” said Dang. “If we want students to have the best experience, we have to keep updating the way we teach.”
Berkeley’s influence extends well beyond campus. Through initiatives like Cal-ICOR and curriculum adoption programs, institutions across California and the country are building their data science offerings using Berkeley’s materials and tools. Improvements made by DSUS faculty and TAs can ripple into classrooms throughout the state and beyond, helping other institutions launch high-quality data science programs more quickly.
While solving problems is exciting, Dang said it’s the end goal that matters to Dang. “We don’t build tools just to build them,” she said. “We build them to help students learn.”
For Data Science Undergraduate Studies inquiries:
brandilyn@berkeley.edu
