WORK
Launching a Bootcamp for Data Scientists
Designing a rigorous program that prepares students for a career in the data-driven economy
The Challenge
Design a curriculum for novice programmers and statisticians to level-up their data science skills and get them job-ready.
IMPACT
Since launching in 2014, the program has expanded to cities across the U.S., as well as online.
The OUTCOME
The Metis Data Science Bootcamp, a 12-week, project-based program that equips students to be successful data scientists and collaborators.
In 2017, IDEO acquired longtime partner Datascope and integrated the company’s data scientists and engineers. Datascope team members aren’t just passionate data scientists; they’re human-centered designers who happen to work with data, as illustrated by successful collaborations like this one:
In today’s job market, there’s no such thing as being "done" with your education. This is particularly true in the fields of data science, machine learning, and AI, where significant advances are made on a monthly basis. To meet this need for more in-depth training, Metis—part of Kaplan—wanted to create a comprehensive data science program for students from all backgrounds. Metis asked Datascope to design the curriculum from scratch, drawing on the company’s deep expertise in the field.
Datascope spoke with practitioners across industries to identify the key skill sets that make them successful: a grasp of design; fluency with data management, coding, and algorithms; and effective communication with coworkers and the wider world. The team crafted a 12-week bootcamp that develops these specific abilities.
In order to prepare students for the work world, the bootcamp is structured less like a class and more like a company, where students are “hired” as data scientists—with projects to complete—starting on day one.
Students collaborate on group projects and presentations in a way that mirrors real workflows, and practice pair programming to learn from one another. Using techniques drawn from design, students are invited to refine their project briefs and continually reevaluate their approach.
Instead of a final exam, at the end of the course, students tackle five real-world, open-ended data science projects inspired by Datascope’s client work. One project involves using data from subway turnstiles to detect patterns in the volume of street traffic; another asks students to use data they scrape from the web to predict a movie’s box office profits using regression analysis.
I loved it; the coursework was a good mix of learning new topics and hands-on work on programming challenges and projects.
Emmanuele Salvati, bootcamp graduate
After students complete the bootcamp, they receive an additional three months of career support, from mock interviews to site visits with potential employers. In addition to designing the curriculum, Datascope taught the first three student cohorts and helped Metis build its teaching team as it expanded the program from New York to San Francisco, Chicago, and Seattle. Metis now offers online courses, evening classes, and corporate trainings in data science.
If you’re going to take a data science bootcamp, it’s best to take one designed by data scientists.