Blog

New Job at McKinsey & Company

boston-skylineThis summer, I spent two months in Burkina Faso, West Africa, setting up our new non-profit, Tractors For Africa. Our tractor arrived in late June (see my previous post), and I stayed through July and into August to train the farmers in proper use and maintenance, re-assemble the equipment we had sent, and setup a sustainable cooperative governance and financial structure.

On my way back to the states, I spent a week in Senegal visiting the agricultural projects of Africare on the ground and learning from their successes and (admittedly few) failures. The Senegalese have had success with peanuts, rice, and fruit crops such as mangoes. Eastern Burkina Faso lacks the rainfall necessary to support mangoes at scale, but increased cultivation of peanuts and other legumes would provide nitrogen to enrich the soil.

After Senegal, I visited Rome, then proceeded to job interviews in Geneva, London, Denver and San Francisco. In the end, I decided to take a hybrid consultant/data scientist role with McKinsey & Company in Boston. I started in September and am currently working on projects as diverse as crop production/price forecasting and drone vision. I work directly with our Deep Learning and Geospatial teams here at McKinsey, and I have already learned a lot from my new colleagues. I am incredibly grateful for the chance to work with the brilliant and pragmatic people here at McKinsey. It is an exciting time to be a consultant as we help our clients quantify uncertainty and make the best possible decisions with modern machine learning and data science techniques.

McKinsey is a rather busy place, but in my spare time, I’m collaborating with professors back at South Dakota State University on two research projects. The first is on grain marketing optimization, and involves stochastic modeling, dynamic programming and control theory. The second is a comparison of the effects of four types of tillage on sorghum yields and soil characteristics in Burkina Faso. You can read more about these projects on my research page.

I was fortunate enough to move in with MIT and Harvard grad students here in Cambridge, MA. One of my roommates is a computer scientist, one is an engineer, and one is a masters students at the Harvard Kennedy School for Public Policy. Through them, I have been invited to presentations and events on the campuses of both schools. The research that is going on within a few blocks of me is incredible. Automation, control, deep learning… the experts are right here, and I get to hang out with them! Truly, it is an exciting time to be alive.

Into the Wild

 

On June 20th, I arrived in Burkina Faso with Tractors for Africa. As this is our first project, its success is crucial for future projects. Another co-founder, Louis Ricard, was with me for the first week in Burkina Faso, and on his last day here (June 25th), the tractor finally arrived! With the help of nearly 50 villagers, we unloaded the container in just 2 hours. People were ecstatic and ready to see the tractor in action, but we had trouble getting it to start (a bit of a downer after raising $30k and sending the tractor 10,000 miles across the globe), and we didn’t have time to fix it before bringing Louis to the airport.

Processed with VSCO

On June 28th, we finally got the tractor started and proceeded to plow the first field. The entire village turned out to watch the affair. One hundred men, women and children watched with joy and amazement as our Farmall 400 plowed the soil to a depth of 16” – much deeper than the 2” that they achieve by hand or with ox-drawn plows. They had never seen a plow before, but they had seen a disk, which was the next thing we tested. The disk only tills to a depth of 6”, but it can go much faster. The villagers actually prefer the disk since it is more familiar, so we proceeded with that. Right now we are tilling land at a rate of 6 hectares per day (due to the small, oddly shaped fields) and we are charging enough to cover gas and repairs. I have trained a tractor operator who is tilling a field as I type. The villagers are already planning to save up for next year so that they afford the gas to till their fields in 2017. I could not be happier!

As for overall project success, we have sent many other pieces of equipment to extend the season. After tillage is finished (around July 15th), we will proceed to row-crop cultivating and spraying for weed control. To my knowledge, these services have never been offered here, but farmers are excited to try them. After harvest, we will transport grain using a tractor-drawn wagon, and during the dry season, we will use the tractor to haul water. In this way, we can use the tractor all year and hope to save enough money to send another within the next year.

I am also planning a study to compare the effects of different types of tillage on crop yield and soil structure. The four tillage types we will test include hand tillage, ox-drawn plow tillage, tractor+disk tillage and tractor+plow tillage. I will post the results on my website after harvest in November.

Commencement on the Farm

Farm Sunrise_edited

This Memorial Day Weekend, my sister Linnea is graduating from high school and is about to begin her career in mathematics at South Dakota State University. She is the likely valedictorian, most valuable female athlete of the year and member of too many organizations to count!

As the diaspora of our family and friends gathers from across the state and country, I am contemplating where my own projects will lead. Louis Ricard and I, two of the three co-founders of Tractors for Africa, will be heading to Burkina Faso during the middle of June. While there, we will train a cooperative of 10 farmers in the proper use and maintenance of farm equipment.

Additionally, I am collaborating with several colleagues and professors on research using machine learning to model biological systems – more details to come. As computer science and applied statistics continue to shape our world, it is a privilege to be a part of it and push the boundaries of what is possible.