Co-PIs:

Marcel Salathé (EPFL)

Problem:

Crowdsourcing data science problems has numerous benefits: multiple approaches to finding solutions, exposing the problem to a large global community of data scientists, etc. But running a successful challenge can be difficult, and collaborating with private USbased platforms is increasingly problematic.

Solution:

crowdAI is an open data science challenge platform coming out of EPFL, and it is completely open source. It is free to use for challenges by researchers and nonprofits, and affordable for industry partners. Solutions submitted to crowdAI are fully reproducible. It can also be used for private (e.g. classroom) challenges.

Impact:

crowdAI has run numerous challenges at high impact conferences such as NIPS, and is used by organizations such as Stanford, Berkley, Cornell, Tübingen, Google Brain, Microsoft, and others.

https://www.crowdai.org