PMI partners with African data scientists

It’s a question that many companies would love to answer—how do you find and leverage talent in parts of the world where you don’t have a strong presence? Over the past few months, Phillip Morris International (PMI) has had the opportunity to discover just how positive a good partnership with the emerging talent across Africa can be.  

Six data scientists from the African Institute for Mathematical Studies (AIMS), recently joined PMI for a 12-week fellowship program to study tobacco-growing areas using satellite imagery. The partnership was the brainchild of Ishango, a unique social enterprise working to increase the opportunities available to talented data scientists all over the continent. 

With the pandemic proving the viability of remote working, they have been helping to improve the opportunities for talented individuals from the continent to work with international companies, in fields such as financial services and healthcare. 

“Our model is to get international companies that have interesting data science projects that our fellows can work on to build skills,” says Eunice Baguma Ball, co-founder of Ishango. “We were thinking: What could really help connect talent in Africa who haven’t traditionally had access to international opportunities, while allowing them to actually remain in their country, contributing to their local economies?”

After Ishango approached PMI, the company agreed to take on six fellows for a 12-week program. They were Bright Aboh from Ghana, Jeanette Nyirahakizimana, Jeannine Uwingabire, and Aimable Ishimwe Manzi from Rwanda, and Nasser Mouliom and Engelbert Tchinde from Cameroon.

The fellows had the chance to develop a generic solution for quantifying the sizes of farmed land, based on the satellite images, independently of what crop was being grown there.

“It’s been a great success,” says Krzysztof Stec, Lead Data Scientist for Operations, at PMI. “It added value. The students proved they were extremely skillful. They were capable of pursuing data science projects with limited supervision. And there were no communication or time-difference issues. They really helped our team.”

The fellows themselves also found the experience educational, enjoyable, and enriching. “I liked the way we were given feedback throughout the project, and the guidance that was provided,” says Jeanette Nyirahakizimana. “I really liked the collaboration which enabled us to achieve our targets.”

“The program has given me more confidence to enter the world of employment,” added Engelbert Tchinde. “I have learnt and developed my technical and communications skills during the fellowship.” He adds that the most valuable part of his experience was to build a classification model which enabled the team to classify bio-geographical regions of the land surface in Malawi and Mozambique.”

The project was all about using satellite images to identify where tobacco farms are located. “We can estimate their area, and from that, extrapolate the yield of tobacco. Based on this, we can make some projections in terms of what can we estimate the price to be,” adds Stec. “That was the primary goal. The secondary was to support the farmers, to see what micronutrients may be missing in that part of the world or how rainfall can affect the yield and generally, see the wellbeing of the crop.”

Creating new opportunities through collaboration

Although this project was only focused on tobacco farms, the potential of the technology is significant, says Baguma Ball. “I really am excited about this project. I’m originally from Uganda and agriculture is such a key aspect of the economy. When Krzysztof says this is something that can be used to identify maize, crop health—that’s a key issue right now. There’s a massive story here.”

Jan Stuebbe, Global Head of Inclusion and Diversity at PMI says he’s extremely proud of the project. “While it may not seem like a big deal, because we're talking about six people, we've never done anything like this before, recruiting from Africa for international projects. We often ran into problems—it’s so difficult in Africa, the distances are great—but Ishango showed that we could do this and work with people on the other side of the world. It’s been a very interesting collaboration between very different organizations, but also very different people in very different locations.”

Stuebbe adds that the potential benefits of this project and collaboration are enormous. “It doesn’t only help our operations because we understand where tobacco is growing, where we can buy it and what the prices could be. It’s also a wonderful engagement tool for African organizations to say to the politicians or regulators that we try to do things that help communities and farmers in Africa. And that increases our standing in those communities and possibly even helps us attract talent in places that we would have never looked at before. The opportunities are huge.”