Real World Data Science DeCal

Lead TA and head of projects: planning pathways across computer vision, NLP, and forecasting, and mentoring students through end-to-end builds.

Berkeley’s Introduction to Real World Data Science DeCal exists to close the gap between problem sets and practice: students build an end-to-end project of their own, from raw data to a working result. I joined as an academic development tutor mentoring those projects, then became lead TA and head of projects.

In that role I planned the project pathways for the course: computer vision, NLP, and time-series forecasting tracks across a range of domains, so that every team had a route that was ambitious enough to be worth a semester and bounded enough to finish. I also oversaw the course tutors while continuing to mentor a team of my own.

Teaching the doing of data science is different from teaching the concepts: the hard conversations are about scope, messy data, and when to abandon an approach. They’re the same conversations I have in consulting and internships, just earlier in someone’s journey.