Editorial Note: This blog was originally published on Atlas AI’s Medium on March 14 shortly after Atlas AI’s inaugural trip to Kenya. On April 2, Atlas AI and the Alliance for a Green Revolution in Africa (AGRA) will sign a memorandum of understanding to bring predictive analytics to smallholder farming across Africa, improving food security and farmer livelihoods. Read the official announcement.
Off to Kenya!
At 5 pm on Friday, Feb 1st, I said goodbye to friends and colleagues at the Wikimedia Foundation, where I’d served as CTO for over two years. Less than 24 hours later, I was on a plane to Nairobi, Kenya to spend my first day at Atlas AI, on the continent we serve and in the country where our founders first started their research.
It was a fitting start to my journey as CEO. In some sense, the technology behind Atlas AI was born in Western Kenya. Co-founder and Stanford professor Marshall Burke, then a graduate student at UC Berkeley, was working on a study of smallholder maize farmers in the region. He observed the challenges — and costs — of collecting accurate information about farmers’ yields and livelihoods. And he could see how better estimates of these outcomes, at scale, might help organizations working with these farmers figure out how to be most helpful.
He brought these insights back to Atlas AI co-founder David Lobell, also a Stanford professor, who had long worked on yield estimation in the U.S. They began to explore how remote sensing-based yield estimates might be adapted for the heterogeneous, small plots cultivated by smallholders in East Africa and around the world. With co-founder Stefano Ermon bringing powerful machine learning algorithms to the problem, the stage was set for Atlas AI. The company was born of the desire to make powerful information and computing tools consistently available to the people who need them most.
Upon my arrival in Nairobi, I met with scientists, students, government officials, and farmers to hear about their experiences and learn what data and insights they need to do their jobs better.
The continent boasts deep talent. I was privileged to meet some incredibly bright minds at the Women in Machine Learning & Data Science meetup and at the U.S. International University-Africa. I saw great passion for doing good and optimism about the role technology can play as long as the technology is applied in the right measure. There’s also an urgent demand for better data. When talking with NGOs and companies, they know that the dearth of timely, high-resolution data is a critical challenge.
Core values of Atlas AI
As we expand our work and presence, our activities will be guided by four core values:
- Atlas AI is an organization that listens. We can’t design the right datasets without input from those who need them for their work.
- Atlas AI is an organization that’s committed to quality above all else. From farmers in the field to heads of NGOs and government officials, everyone I met wanted to be confident that our estimates were as accurate as possible. I’m proud to say that our work is driven by peer review research published in the highest quality journals. We want to partner with people who expect the highest quality and can tell when that’s missing.
- Atlas AI is an organization that will be on the cutting edge of machine learning. We want to cut the time-to-market in half for algorithms that can support the SDGs.
- Atlas AI is an organization that complements the work of others. We’re built on openness and the commitment to build long-term value. We know we can’t do it all ourselves, and we will work with partners on this journey.
These are the values that our founders and partner, The Rockefeller Foundation, are committed to, and they are a big part of why I chose to join as CEO.
What to expect from Atlas AI in the next few months
This quarter, we’re launching a beta user program to allow our data to be road-tested by the folks who are on the ground: leading NGOs working to improve agronomic practices, world-class consultancies advising governments, and research and evaluation teams examining new methods.
We’re also testing new verticals. We heard great demand for better data on transportation and infrastructure. These are areas where we think we can contribute, and we’ll be releasing data sets later this quarter.
We’re building our team, our partnerships, and our network.
This is only the beginning of our journey, but we’re excited to share what we can do today and build for tomorrow. If you need faster, higher resolution measurement of economic and agriculture activity, visit us at www.atlasai.co or reach out directly at email@example.com.