Big Data, Big Impact
April 16, 2014

At our March Meetup, our current group of DataCorps volunteers shared the results of the data science projects they tackled alongside four inspiring organizations. They talked about the challenges: freeing data from non-delimited text files, navigating proprietary software, and harnessing visual data from satellite imagery-- and the impact: saving, protecting and improving the quality of life for people around the globe.

The DataCorps is a pro bono service program for talented and expert data scientists who offer their time pro bono to serve non-profits with great data, clear questions and the desire to use insights from a data science project to maximize their impact.  

As DataKind volunteer Katya Vasilaky put it: "This group of DataCorps members tackled two of the hardest issues that NGOs face: measuring the impact of their programs (Medic Mobile and The Mission Continues), and figuring out how data could maximize each organization’s impact (GiveDirectly and Thorn)."

Cat Miller and Will High worked with Medic Mobile -- a Skoll-award winning non-profit that fights to provide more and better access to public health in the developing world. Medic Mobile wanted to measure the impact of their health interventions and thought that John’s Hopkins University’s Lived Saves Tool (LiST) might be the right tool to use for the data they'd collected. Cat and Will tested this theory and built Medic Mobile a creative guide and interface to help them know when and how to best use the LiST tool-- and when their data called for other methods of impact measurement.

GiveDirectly approached us with a huge question: could big data help them identify the best villages in Kenya to launch their innovative cash-transfer programs to combat poverty?  GiveDirectly knew that the number of thatched roofs in a village was a key predictor of income and poverty levels in Kenya and many other parts of East Africa. When a family has to worry about fixing a roof after a rainstorm, it can be hard to get ahead.  

Kush Varshney and Brian Abelson jumped in to help. They developed an impressive machine learning algorithm to sort through satellite imagery and identify villages with a significant number of thatched roofs. But they didn't do it alone. In order to train the algorithm, they needed to label 1500 satellite images by hand.

They asked the help of DataKind's NYC Meetup, thinking they'd be able to make a dent in the task. Three mornings later, Brian woke up to discover that ten volunteers had finished the entire job and labeled every last image! Thanks to the hard work of the DataCorps team, GiveDirectly is now saving 100 days of person-effort to identify the villages in Kenya where their cash transfers can have the biggest impact! 

Seth Feaster, Sonali Murarka, Sonia Mehta and Burton DeWilde teamed up with Thorn to deal with some difficult data and even more difficult problems. Thorn's mission is to put a stop to child sex trafficking. Their data collected aspects of behavior on websites where trafficking was known to occur. With grace and deep sensitivity, the DataCorps team found a way to turn seven million records (in a non-delimited text file!) into usable insights that Thorn can now share with law enforcement and partner NGOs working to protect children.   

The Mission Continues is also working to protect and serve, by providing vets returning home from Iraq and Afghanistan with opportunities to volunteer in their communities and readjust to civilian life. They collect surveys and demographics data from each of their program participants. Alex Vaughn, Chris Tynan and Tushar combined their skills in quantitative and text analysis to see if they could identify the factors that help a veteran to succeed in The Mission Continues program, and ways in which The Mission Continues could tailor their programs to better support the growing numbers of post 9-11 veterans.

Regional Director and Iraq veteran Aaron Scheinberg was on hand to share just how much the DataCorps' work meant to his organization: 

The Mission Continues is thrilled with our experience with Datakind and the skilled data scientists that joined us.  They were not only able to help us analyze existing surveys to improve our programs for Post-9/11 veterans, but also became trusted advisors in how to improve our data collection and research strategy. The team was a pleasure to work with and provided much needed capacity and insight during a critical growth period.

Our DataCorps service program is a crucial part of how we provide support to the social sector. Not only do these talented volunteers help organizations use data to solve big problems, they're helping to promote a data-informed culture within the nonprofit sector.  We can't thank our volunteers or our partner organizations enough for their commitment to use data in the service of humanity. Please leave a comment below if you'd like to join us in thanking these amazing people for their time and dedication!

We are currently recruiting skilled data scientists for the 2014 DataCorps. If you're interested in being considered or learning more, make sure you've signed up on our site!

Check out photos from the DataCorps presentations >>

Read more posts
April 12, 2021
Powering Public Data for Communities: Highlights from Virtual DataDive®️ Report Back
We recently hosted our DataDive Report Back, keeping with the theme of Trust, Transparency, Togetherness: Powering Public Data for Communities, and we’re honored and humbled by the level of excitement...
Read full story
November 07, 2019
Applying AI to Societal Challenges in Cities & Communities: Highlights from DataKind’s Virtual DataDive With Microsoft
We can’t believe it’s been a month since our first-ever virtual DataDive. We’re still feeling the energy and enthusiasm from that day. More than 100 DataKind volunteers from around the world, hailing from...
Read full story
June 20, 2018
Using Satellite Imagery to Generate Awareness and Funds for Refugees
At a 2017 DataDive presented by USA for UNHCR, #GivingTuesday and the Bill & Melinda Gates Foundation, a team of DataKind volunteers led by Evelyn Lang, Ravin Pierre, and Friederike Schüür worked alongside...
Read full story
Blog Archive