“Zindi and Local Ocean Conservation (LOC) share the belief that we live in an interconnected world filled with talent; in our case the talent and technical expertise we needed was not in house. A Zindi campaign enables us to get solutions we wouldn't otherwise have been able to get,” says Justin Beswick, CEO of Kenyan conservation organisation Local Ocean Conservation. “It’s really valuable to step out of the conservation space and tap into other industries, sectors and skills, such as Zindi's data science community.”
LOC is a small organisation based in Watamu, Kenya that has been protecting and rehabilitating turtles for more than 20 years. What started in 1997 with local residents wanting to protect endangered sea turtles has expanded into a successful marine conservation organization, using turtles as indicator species to monitor and protect a number of marine and coastal environments. One of the keys to their success, Beswick explains, is in the data.
“The turtle programme has been run well in terms of international protocol and best practice for conservation and data collection. Our data collection has been good enough that it is now one of the few datasets on turtles in the region that can be used in academia. We have more than 20 years of properly collected data on turtles - this can give us key insights into green and hawksbill turtles, the green turtle is endangered and the hawksbill is a critically endangered species.”
As part of a recent digital transformation push at LOC, they wanted to clean up the dataset so that it could be put to better use in monitoring impact, obtaining insights and driving business decisions. And so LOC turned to Zindi, for the Sea Turtle Rescue: Error Detection Challenge.
“We came to Zindi to fast-track data cleaning and verification for our bycatch programme data, so the data could be used for research and to inform management on the success and impact of interventions,” Beswick explains. “Machine learning greatly reduced the time and resources we had to allocate to the data cleaning process. Zindi helped us find machine learning solutions and implement them on our database, creating outputs we can use to inform our decision-making.”
LOC were so happy with the result that the collaboration has continued with a Turtle Rescue Forecasting Challenge that’s about to wrap up on Zindi. There is also another challenge planned for later this year that will call on Zindians to build an AI-driven turtle facial recognition app.
But the payback for LOC goes beyond just machine learning solutions for their organisational challenges. Beswick says the competition and the collaboration with Zindi attracted the attention of technology-focused supporters and funders, in particular Microsoft.
“Through Zindi we were put in touch with Microsoft, and we’re now part of their 4Afrika programme. Machine learning and what we’re trying to do with Zindi also got a lot of people in the fundraising world interested in us, as it showcased our efforts at the forefront of using technology for conservation.”
For Beswick, what we’ve done together is about much more than getting the right answers out of the data:
“Whenever we reach out to people working beyond conservation, whether software engineers or data scientists, we see that the opportunity to apply technical skills to conservation is enjoyable and empowering and people really want to help. They just need to be enabled to help.
“There is a deep pool of data science talent coming out of the continent, but African skills and technology are constantly overlooked or disregarded. This was true a decade ago, and it persists despite the many success stories that have come out of African tech startup communities. But we're continuing to debunk that myth.”