Omdena Talent Helps Catholic Relief Services Create and Scale AI Algorithms That Address Poverty

Interviewee: Kathryn M. Clifton, PhD

What are the most relevant recent success stories from CRS?

Hiring Christian Fokoua and Bruno Paixao from Omdena has helped us scale ML algorithms to multiple countries that help us answer important questions like who is not showing up to services, what services are working the best to reduce poverty, and who is most at risk for food insecurity.  We went from 1 to 2 Omdena staff because country programs have found value in this work and we have had an increase in requests and willingness to pay for support. The fact that both Bruno and Christian are from countries that have similar problems has helped them understand the problems on the ground and deliver helpful solutions. Also, the staff that Omdena provides also cares deeply about using ML to make the world a better place and that has saved me a lot of time in recruitment. I no longer need to give employment tests to test ML knowledge and spend a lot of time understanding if our values align.  

What are your focus areas in terms of milestones to accomplish in the next 12 months? 

Create new ML algorithms to match skills with employment for Gaza and the West Bank. Scale our ML algorithms for who is now showing up and what approaches are working best for 5 programs in Water Sanitation and Health. Scale our ML algorithms to 3 food security programs. Also, to create a new algorithm on who is most at risk for water insecurity.  

What frustrations did you/ your team experience before working with Omdena?

I spent a lot of time hiring to have people take a better offer at the final negotiation phase. It was hard to find ML skills who were invested in our mission. We had ML work paused because of staffing. We were often the last pick in ML challenges where we were up against google and amazon because students were trying to get jobs with them.  

What were the greatest challenges you felt during normal operations?

We would get AI results from partners that were static. It made it hard for programs to know how significant results are changing on the ground near real-time and what to focus on. Omdena staff were able to make algorithms to continually update and run and display those results on dashboards. It turned AI from a research concept to an operational tool.  

How did working with Omdena Top Talent AI teams provide better results?

For the first time, I was able to work with people that were not only invested in how we can use ML to solve problems around poverty but also were interested in a long-term career with us. I was able to find people with humility who would work as team players and help build the skills of my non-ML staff. This attitude difference resonates in everything that they do. I have to remind them to take vacation leave,  they are passionate and dedicated. This makes my job an absolute joy, I am no longer managing egos but managing the best way to solve a problem and scale the solution and they provide essential feedback. We work together as a team. My other team members have thoroughly enjoyed having them because they have grown as a result.

If and how has the work culture changed for the better since working with Omdena?

Often in advanced concepts ego can be brought to the table and people can use that as a barrier to entry since knowing something everyone else doesn’t know gives people a position of power.  As a female scientist, it is very important for me to have a culture of inclusiveness and have a pathway for anyone willing and interested to learn. There are no dumb questions, just dumb answers is an important motto for me. Christian and Bruno embody a growth mindset that is open to all.  As a result, one of my non-ML staff is now running ML algorithms with coaching from Christian, and another is running NLP also with coaching from Christian. We have tried many courses but Christian’s coaching has given us the best results.