Fellows typically come from Computer Science, Statistics, Math, Physical Sciences, Social Sciences, and Policy backgrounds. They are mostly grad students (and some advanced undergrads and post-docs). You can read about the applicants from 2015 here.
We typically look for a mix of skills, experiences, and backgrounds. We believe that it requires a diverse set of people to solve the problems we tackle and we encourage everyone passionate about making a social impact with data science to apply. We do, however, expect a base level of computational and quantitative skills. We can’t be everything to everyone so we’ve decided to focus this program for people who have at least some computational and data analysis experience.
Typical fellows have reasonable proficiency in at least one programming language (we typically use Python) and some experience in statistics and data analysis. We are looking for a portfolio of fellows so we can create balanced cross-disciplinary teams. Typical teams consist of computer scientists, statisticians, and social scientists. Teams have access to dedicated hands-on technical mentors and project managers, as well as a larger pool of expertise within the fellowship.
Read more about what we’re looking for in the FAQs as well as our blog posts.
Mentors are a critical part of the fellowship. Mentor positions are full-time and (modestly) paid. Ideal mentors have a strong data science background (typically PhDs) with significant industry experience. We don’t expect mentors to have worked on problems in social issues but we do want them to have hands-on experience working on real-world problems. Each mentor will be in charge of two teams, and will work closely with the teams on the projects on problem formulation, algorithms, coding, evaluation, etc.
Project Managers ensure that the project is moving forward, the team is functioning, as well as manage the relationship with the Project Partner Organization.
We look for project partner organizations that can provide us with projects that:
- Have social impact
- Can be solved using data science