A Tale of Data, Communities and Opportunities

Ulrich Mabou | Tuesday, April 16th, 2019

How MEST is supporting the growth of a data science & AI community in Ghana

This post was written by Ulrich Mabou, an Entrepreneur-in-Training in MEST Africa’s 2019 cohort.

Back in September last year, I wrote here about the importance for African nations to develop the policies, infrastructure, ecosystems and skills necessary to put us in a better position to embrace exponential technologies and solve our local problems ourselves. I am glad to realize that a lot has happened over the last six months and that Africa is gearing itself up to exploit the full potential of the fourth industrial revolution.

Personally, I have tried to commit more of my time towards gathering skills in data science, machine learning and related fields. Thanks to Robert John and MEST Africa’s management, some of us at MEST have been doing some very interesting things on the Google Cloud Platform (GCP) lately. Here are some of the quests (series of tutorials or ‘labs’ on a specific topic) I have completed so far.

The journey has not been easy, but with the proliferation of freely available learning resources, the support from the MEST community and the hope of great future opportunities, I am slowly but surely equipping myself

In this piece, I will share three key things I have recently been involved with (actively or from far) on my data science journey and within the MEST community. They can be summarized as: datacommunities and opportunities.

Data: Data Upload Challenge & Hackathon

EITs hard at work at the data upload hackathon. Photo from Mobile Web Ghana.

Two weeks ago, some of my fellow EITs (Entrepreneurs-in-Training at MEST Africa) and I took part in the first edition of the Data Upload Challenge & Hackathon. Organized by Mobile Web Ghana under the leadership of MEST alumni Florence Toffa, and supported by the IFC — International Finance Corporation, the event challenged civic technologists, statisticians, journalists, software developers, and data scientists to develop data visualizations and innovative applications using datasets supplied by various ministries and governmental agencies.

Jamila speaking at the data upload hackathon

Government Ministries, Departments, and Agencies uploaded some of their datasets to the Ghana Open Data Portal. We participated to support the Ghana Open Data Initiativebecause we strongly believe such initiatives have the potential to positively impact problem-solving and opportunity creation in Africa. With our very own global open data advocateJamila Farouk Jawulaparticipating, we were eleven EITs in total and we split ourselves into two teams.

Team Simplify

Our team (Simplify) used datasets from the Registrar General’s Department and theGhana Investment Promotion Centre to work on a portal with visualizations that help business owners know which sectors are booming, where there is Foreign Direct Investment (FDI) and businesses currently registered in those sectors. Our hope was that insights gathered from our analysis could help the government and agencies prioritise their budgets and enforce policies.

We were also hoping to increase their productivity by cutting the time they spend attending to recurring calls and questions from the public. With up to 17 teams pitching at the end, the competition was very fierce. While our team didn’t win, we learnt a lot, mostly on the importance of encouraging and supporting Open Data Initiatives.

Team AgroIQ (second runner-up at the data upload hackathon)

Our second team (AgroIQ) came in third with a machine learning model to predict future prices and yields of various agricultural commodities. They used datasets from the Ministry of Food and Agriculture(rainfall, historical prices and yields) and the Ghana Statistical Service (demographics), coupled with some datasets from the World Bank to predict future prices and yields for different types of crops, including maize, millet, sorghum, cassava, cocoyam, plantain, groundnuts and beans. They managed to build a pretty good model and an intuitive web application in two days; the core of their value proposition was on insights, trends analysis and prediction capabilities.

Communities: Accra Practical Data Community

Attendees at the first meetup of the Accra Practical Data Community

Under the leadership of senior faculty Michael Leventhal, we started the Accra Artificial Intelligence Community back in August 2018. It is a community-based learning group around artificial intelligence and machine learning focused on enterprises and applications targeting Africa. One of our goals was to learn about artificial intelligence together, so each week a topic was proposed, and one or more participants would volunteer to present a tutorial on the topic and lead a discussion.

After a great start, people started losing interest, which caused us to realize that we spent too much time discussing theory and applications, instead of learning the hard skills required to become a data scientist or machine learning professional.

As a result, we started the Accra Practical Data Community (PDC) with the aim of building a sustainable practical data community in data science, artificial intelligence, machine learning and related fields. Our first meetup was held on the first Saturday of April 2019.

Feel free to join our meetup group here and register for one of the upcoming sessions; you can also join our Slack workspace here.

Jamila leading a session on Python programming during a PDC meetup

The first meetup included a practice session with Funsho Olaniyi taking us through an introduction to Python and Pandas using the same datasets his team (AgroIQ above) explored during the data upload hackathon. You can check out what we covered together on this repo. We ended with a presentation on Data Ethics by Abdulmalik Ibrahim followed by a brief discussion. We learned that the six principles of data ethics are: ownership, transaction transparency, consent, privacy, currency and openness.

Opportunities: DataHack4FI Training Program & Competition, and more…

Some of our founding members discussing with the CTO of Complete Farmer at a recent DataHack4FI event

Over the last four months, some EITs have been seriously crunching some data as part of the Microsoft Professional Program in Data Science. They enrolled for the DataHack4FI challenge, an initiative of Insight2Impact that aims to pair data enthusiasts and emerging tech companies in Africa. After completing some course modules, six of them made it to the next step: the capstone project.

Ashwin Ravichandran, Director of Portfolio Support at MEST Africa, introducing the DataHack4FI Capstone Project

They are each going to be paired with a participating tech company to develop solutions that address local financial and economic inclusion-related challenges. DataHack4FI also serves as a platform for each participating team to showcase the solution they develop to potential investors, partners and clients. Now in its third season, DataHack4Fi accepted applications from six companies in Ghana this year, including MEST portfolio companies Complete Farmer, Growth Factor Technologies with their product Nvoicia, and Damansah.

They will all have access to mentorship from data science and business experts during the competition. The teams from each of the countries are competing for US$ 5,000 in seed capital. In addition, in-country winners and runners-up will be sponsored to attend the Grand Finale where teams from the seven countries will compete for US$ 25,000 in seed capital. The paired data enthusiasts will all receive Microsoft’s certifications in Data Science after completing the capstone project with their respective teams.

The future looks bright for Africa!

Young Africans want to learn data science, AI and machine learning

Referring back to my article last year, I reported that only Kenya and Tunisiahad taken steps towards developing a national AI plan. It is interesting to note that many African nations have made a tremendous effort over the last couple of months to change that. Not later than last week, the South African President Cyril Ramaphosa has appointed members to the newly created Presidential Commission on the Fourth Industrial Revolution (4IR) in an effort to strategically position the country. The task of the commission, which will be chaired by the president, is to identify “relevant policies, strategies and action plans that will position South Africa as a competitive global player”.

“The commission will assist the government in taking advantage of the opportunities presented by the digital industrial revolution” — In a statement from the Presidency of the Republic of South Africa on Tuesday, April 09 2019.

Moustapha Cisse at the opening ceremony of Google AI Accra (April 10 2019)

Special mentions also go to the governments of Rwanda and Ghana for continuously creating policies that facilitate the learning and applications of new technologies, as well as supporting the development of strong research and entrepreneurial ecosystems in their countries.

That is one of the reasons why the MEST training programwas initially set up in Accra more than ten years ago. In more recent news, Google officially opened its AI Research Centre in Accra last week. With a starting research team made up of a group of diverse individuals from different backgrounds and countries, the centre aims to collaborate with local universities, research organisations and policy makers in Ghana and across Africa to deploy AI in solving challenges in the healthcare sector, agriculture, education, and other sectors.

As expected, the demand for data science and AI talent is also skyrocketing, both globally and locally. In Africa, CodeLn, a MEST portfolio startup that facilitates testing and recruitment of tech talent, regularly receives enquiries for data scientists and machine learning engineers. According to LinkedIn 2018 Emerging Jobs Report, “Artificial Intelligence (AI) is here to stay.”

Ulrich is currently an entrepreneur-in-training at MEST Africa. He is a data science enthusiast who is very passionate about the possibilities of combining IoT and machine learning to solve problems and unlock potential in Africa. He writes about tech, business, entrepreneurship and productivity. Feel free to connect with him on LinkedIn or follow him on Twitter.