The Bayes Impact Mission

Solving ambitious social challenges using data science

“The best minds of my generation are thinking about how to make people click ads. That sucks.” — Jeff Hammerbacher, Chief Scientist at Cloudera

Over the past decade, we have all benefited from a better use of data: self-driving cars, personalized movie recommendations, smarter drug discovery — the list goes on. The underlying driver behind all of this is data science, an integration of statistics, computer science, math, and domain knowledge to extract knowledge from data. Amazon uses it to recommend items to add to your cart, Google and Facebook use it to show you the most relevant ads, and Uber uses it to dispatch drivers to your location faster.

With data science, we’re able to build increasingly complex systems at scale that not only provide insights but also are able to offer actionable predictions. It is an incredibly formidable tool that allows us to leverage the work of a few individuals to affect the lives of millions. If the tech sector is using data science to break open consumer and enterprise markets, we have to ask ourselves: why haven’t we seen the same disruption in solving social challenges?

The answer to that is multi-faceted:

1. First and foremost, integrating data-driven solutions to high impact problems is hard. Solving these types of problems means convincing multiple stakeholders, governments, and private entities to change what they’re doing. It involves deeply transforming existing processes as well as working with messy (sometimes non-existent) data. Achieving impact requires a deep engagement model, deploying teams with the same level of focus and determination that successful startups need to have in disrupting markets.
2. Second, getting smart people to work on social impact problems is hard. Too many talented people go into finance and tech startups — and why shouldn’t they? The social sector has a reputation for paying less and moving slower, and risk taking on new approaches is not encouraged by taxpayers nor the existing philanthropy model. If we’re going to solve the really tough problems — prison overcrowding, homelessness, decreased competitiveness of our education systems, and so on — we need to convince the very best talent to shift their time away from selling ads and on to solving impactful problems.
3. Finally, learning how best to utilize data science from scratch is hard. It’s not a secret that the social sector lags behind industry and academia in technology. The right data is not always captured and, when it is, not used to its full potential. While some civic and nonprofit organizations are making great strides in this area, many will struggle initially with structuring data science problems, managing a data science team, and integrating solutions into underlying processes.

At Bayes Impact, our mission is to fix this. Our nonprofit deploys teams of talented data scientists to work on the world’s toughest social impact challenges — from reducing prison recidivism by building better risk assessment models to making micro-finance more economically viable through fraud detection to looking for early indicators of neurodegenerative disease using machine learning techniques. Solving high impact problems requires focus, so all of our data scientists are full-time fellows that dedicate 100% of their time on a single social impact challenge for 6-12 months. To attract the best talent, we work on only the most interesting, high impact projects and provide mentorship from data science leaders in industry and academia. And to bridge the gap between the social sector and data scientists, we manage the projects the entire way, from scoping and defining problems through to execution and long-term sustainability.

We already have some great people working on greater problems. If you’re a talented data scientist looking to do more with your skills or a social good organization looking to do more with your data, reach out to us. The social sector doesn’t have to be left behind in the data science revolution—let’s solve the world’s toughest problems together.

www.bayesimpact.org