‘Don’t let you first hackathon be your last’: tips and trick for UmojaHack Africa 2023, from 2022 winner Lawrence Moruye
With UmojaHack Africa 2023 just days away, we caught up with Lawrence Moruye, winner of UmojaHack Africa 2022, to get some tips and tricks on how to succeed in this year’s event. If you’re interested in participating in this year’s event, check out umojahack.africa to sign up.
Can you please introduce yourself? Tell me a little bit more about yourself, where you’re from and what you’re currently doing.
Lawrence Moruye: My name is Lawrence Moruye, I am from Nairobi Kenya. I did my bachelor’s degree in Mathematics and Computer Science from Multimedia University of Kenya, and I graduated in 2020 with First Class Honours. And then I got a scholarship sponsored by Google and Facebook for a Masters in Machine Learning at AIMS (Africa Institute for Mathematical Sciences) in Senegal. That began in January 2021, and I will be graduating this year.
Prior to that I also did an internship in data science with a company called Qhala, based in Nairobi. I have also worked as a data science instructor for Huru School for a couple of months. And then more recently I started working with a pan-African e-commerce company called Jumia, where I’m currently working as a data scientist. I just got an offer of full-time employment with them.
Africa's inter-university machine learning hackathon - Umojahack Africa 2023
3000 Students, 300 Universities, 30 Countries. Join Zindi for Africa's biggest inter-university machine learning…
Tell me about your data science journey up to this point.
In my second year, I was trying to look for a career that combines both mathematics and computer science. I found data science as a career on Google, and I started doing courses on EdX, Udemy and Udacity.
A few months later, Zindi actually launched at a hackathon in Nairobi. I couldn’t make it to the event, but I saw Zindi as the perfect opportunity to boost my skills. I joined my first competition on Zindi, but I really didn’t do well. But once the competition was over, then I saw a blog by Mohamed Jedidi, who had won the challenge. I learned some feature engineering and a few tricks from him, and I kept on participating in upcoming competitions on Zindi.
A few months later, Zindi launched the Mtoto News competition, I’d already participated in a couple of competitions in between, kept on improving my skills and then I managed to win first position in that particular competition. So that was one of the biggest motivations for me — I spent a couple of months trying to learn and also trying to put my skills in practice, and I managed to get first position. I remember during that time, I participated in almost every competition that was launched. And if you try every competition, you will have a chance to put new skills into practice, even if you are not winning. You have competitions on NLP, on computer vision, tabular data etc.
Mtoto News Childline Kenya Call Volume Prediction Challenge
You may participate in this competition as an individual or in a team of up to four people. When creating a team, the…
During 2020 the country was locked down, university was suspended and there was an announcement that we have to go home. I didn’t know when I would return to university, so I reached out to Celina Lee (CEO of Zindi) to see if there were any opportunities for internships. I had participated in a Zindi competition which was on a mental health challenge, and Zindi linked me with Qhala who was responsible for deploying the model. So, I went to work with them in deploying the models I had built. I worked with them for six months, and then immediately after graduating I got my scholarship for a Masters programme in Senegal.
Once I was done with my masters coursework, I got an opportunity to start working with Jumia as a data scientist while working on my graduation research project.. So that’s where I’m still working currently. They have some interesting topics that we’re working on, mostly in the field of machine learning and data science, and it’s an opportunity to test my skills and apply some things. It’s a space that I am interested in working in.
Tell me about UmojaHack Africa 2022 — How did you go about approaching the problem that you tackled? And what was the experience of winning the competition? Did it have any impact on your lives?
For last year’s hackathon, I didn’t set aside time or do some specific preparations as I have for the previous editions in 2020 and 2021. If you have been participating in Zindi for one or two years, you are already familiar with machine learning challenges, that’s great preparation. It gives you the confidence to approach any problem.
The only thing I did is to have my pipeline and working environment ready to save time, because the competition is happening in a limited time and probably you have a couple of ideas to try. You can plug in your data and quickly get to a baseline and then from there, apply a few things, try different approaches. When I approach a competition, I try to get a baseline model using almost all the features that are already provided, and then do a little bit of EDA.
Exploring your data is the first step to building a great machine learning model: a tutorial from…
We live in a world where huge volumes of data are generated every single day, from our smartphones to what we search on…
So that’s actually what I did for UHA22. I used all the data provided and my first submission gave me a good position on the leaderboard. I went back, I did a little bit of EDA to see what the data looks like, and made my second submission. I came back to the first position and that was it.
Generally speaking, one of the benefits of doing well in competitions is that people will reach out to you privately, for work or for mentorship, you become well-known in the community. I can say specifically for 2022, the beneficial thing was the cash prizes, of course. This was like a motivation, a recognition of what you’ve already done.
What are your plans for the future?
My overall career goal is to become a leading expert in the field of machine learning and also contribute as well to research that advances the field and also benefits society.
I believe Africa represents a unique opportunity for machine learning research and there’s a tremendous potential for technology to bring impact, specifically in a few areas, for example, in agriculture. And I believe I will be able to contribute to those impacts and benefit society using data-driven approaches.
For a lot of people, this will be the first UmojaHack Africa 2023 will be their first hackathon, maybe even their first attempt at machine learning. It’s obviously very daunting and scary. What would be your advice? How should they get started?
As somebody new to data science wanting to build your career in this field, the best advice I can give is to really build the fundamentals in data science — the basic maths that you need, programming, and a few visualisation or data exploration skills. There are new tools that are coming up every day, but these are the core fundamentals and they’ll always stay the same.
Try to develop communication and collaboration skills as well, because this is a multi-disciplinary field where you’ll be working with data engineers, software engineers and product teams at different times, and the key is to be as effective as possible in communicating with each of the teams.
And for those that will be taking part for UHA2023 as their first hackathon, try as much as possible to pick up and focus on just one competition. Even if you don’t manage to get to a prized position, at least you will learn something. And my advice for you all is: don’t let your first hackathon be your last hackathon! After UmojaHack Africa 2023, participating in several hackathons will allow you to master several different skills that will be very helpful in your future careers.
If you’re interested in participating in this year’s event, check out umojahack.africa to sign up.