What Is DrivenData Competitions?

Updated

DrivenData is a platform that runs online machine learning competitions with a social impact focus, connecting data scientists with real-world problems in areas like international development, health, education, research and conservation, and public services. It is best suited to data scientists who want to apply their modeling skills to mission-driven challenges, and to organizations that have hard problems but lack access to data science talent and tools.

What DrivenData actually does

DrivenData describes itself as working "at the intersection of data science and social impact." Its activity falls into three main areas:

  • Online modeling competitions. It finds real-world questions where data science can have positive social impact, then runs competitions for data scientists to develop the best models to solve them.
  • Direct work with mission-driven organizations. Beyond competitions, it works directly with organizations to drive change through data science and engineering.
  • Open-source projects and shared solutions. It maintains open-source projects for the data science community and has published prize-winning solutions from past competitions openly on GitHub so anyone can learn from and build on them.

Who it is for

If you are… What DrivenData offers
A data scientist wanting socially meaningful work Competitions built around real problems, plus open solutions to study
A mission-driven organization A route to data science capability you may not have in-house
A learner or developer Open-source projects and publicly shared winning solutions

The platform states it has worked with more than 80 organizations across 150+ projects, many made possible by the DrivenData community.

Examples of the work

The site lists several projects that show the range of problems:

  • Public health: With Yelp, Harvard University, and the City of Boston, DrivenData combined Yelp data with Boston's open data on past inspections to predict public health risks at restaurants. The algorithmic approach discovered 25% more violations.
  • Conservation: With the Max Planck Institute for Evolutionary Anthropology and the Arcus Foundation, it ran a global algorithm development challenge to identify wildlife in video footage automatically and at scale, then built an open-source application from the winning solution (the Zamba project).
  • Development: With IDEO.org, the Bill & Melinda Gates Foundation, and Airtel, it analyzed millions of mobile money records to uncover behavioral patterns, combining those insights with human-centered design to shape new approaches to delivering mobile money to low-income populations in Tanzania.

How to get started

If you want to participate, the entry point is the competitions section of the site, where you can browse open challenges and see what each one involves. If you represent an organization with a problem that data could help solve, the site points to its process description and project examples as a way to understand how a collaboration works.

One practical note: the site mentions a "subscribe" option, but the available material does not specify pricing, account requirements, or access limits for competitions or services, so check the individual competition or contact page for those details.