Title: Junior data analyst, SafeBoda
Location: Nairobi, Kenya
Education: BS, Jomo Kenyatta University
Favorite ML area: Natural Language Processing
Favorite ML researchers: Andrew Ng
Kennedy Kamande Wangari got his first taste of AI just two years ago in Kenya. Since then, he has been sprinting towards a top-tier career by taking numerous online courses, working junior-level jobs, helping to organize his local AI community, and considering a startup. Next step: Earning a graduate degree. Below, Kennedy shares lessons he learned from starting out at a breakneck pace and explains how he balances professional obligations with personal life.
Tell us a bit about yourself.
I’m a junior data scientist at SafeBoda, a mobile platform based in Kenya that provides urban transportation for Africa’s booming population. Mainly my work focuses on marketing analytics. Previously, I worked as a junior data scientist at Conflux.ai.
I am an ambassador for several tech communities, among them deeplearning.ai, the AWS Educates Cloud, and the Google Developer Student Club at Jomo Kenyatta University. Beyond that, I’ve been involved in AI for Good.
In October, I will join the University of Michigan. I aim to get a master’s degree in applied data science.
How did you get interested in AI?
I didn’t have access to electronics when I was very young. It wasn’t until I was in fifth grade that I first saw a mobile device. This really sparked an interest in me and led me to take up computer studies in high school. When I went to university, I started to find myself in more and more data-related courses, and realized I was getting deep into business intelligence, data warehousing, and data mining. Getting an internship with a local bank as a business intelligence analyst was the pivotal point in my life. It made me realize my passion for working with data and using it to solve customer’s problems. I found my first data science mentor at the bank.
In addition to your formal education, you’ve taken quite a few online courses. Do you have a strategy for how you piece together various courses?
Before taking any course, whether it’s online, part of a boot camp, or the graduate program, I always consider whether the program will help me achieve my objectives. I take courses that will help me continue building up my skills.
When I was younger, I envisioned being a doctor or taking up a health care-related role. So when the AI for Medicine specialization came up, I felt a burning interest to understand more about the intersection of these two fields.
What have been the biggest obstacles for you when it comes to learning AI?
One of the challenges has been getting to understand the skills in the courses. But the main challenge comes afterward, when you implement that knowledge in real-world projects. How do you structure projects? How do you communicate your goals to both technical and non-technical people? How do you handle various peoples’ expectations? How do you move from ideation to a tangible prototype?
Those skills were hard to gain from taking courses alone. I was lucky to have strong mentors in marketing analytics who taught me how to prove my competencies.
It has been quite a challenge keeping pace with the relevant skills for all the new trends in the field, as well as my own project ideas. Also, learning to say no is something I’m still working on.
What advice do you have for people who are looking for mentors?
The first step is defining what you want out of your career path. Then you can search for mentors who have followed similar paths. You won’t become like them exactly, but it will help you get on the right path. You can start by looking in your own professional network. Consider a mentor who is familiar with your current role. This person will be able to share advice on new projects, certifications, and training you need to get ahead, and how to manage office politics.
What is it like to be an ambassador for deeplearning.ai?
So now we come to the community that I love most: deeplearning.ai! I started working with deepLearning.ai because I wanted to play my part to pave the path to innovation and power socio-economic development. I believe in the mission to create a workforce capable of building an AI-powered future.
We had our first event in February. We now have six ambassadors and over 300 deep learners in Kenya, Rwanda, and India through physical and virtual meetups. Six hundred people attended our last event. We have negotiated valuable partnerships and collaborations with various AI hubs, tech communities, and universities.
I’ve grown exponentially and become a better negotiator, administrator, speaker, and event planner. I’ve interacted with esteemed data professionals and gained mentors who have guided my work. I’ve even been offered speaking and writing roles apart from actual career opportunities. I’ve been invited to attend ICML 2020 as a volunteer and NeurIPS 2020 as a super-meetup organizer. I attribute our community’s success to the amazing Kenyan Event Ambassador team and their tireless passion and dedication for AI.
What will you focus on in your graduate studies?
My main interests have been around natural language understanding, specifically automatic text summarization and information extraction. I’ve been tackling intent identification and recognition from natural language utterances, and understanding of language and its unspoken components. NLP is the most transformational technology in AI. The future looks so promising in this field, and I would love to be part of that transformation. I have some crazy ideas for what I want to do with the knowledge and advanced skills I will gain from the program.
Can you tell me more about your crazy ideas?
My long term goal is to have a startup based around an AI core technology, so for now it’s hard for me to share more details about what we are building.
How can you juggle all these things at once?
I have struggled with this in the past, but now I have activities in prioritized order on my to-do list. I’m trying to balance my responsibilities with work, and those with my various communities, and also have a social and personal life. It has been chaotic at times, but I build habits so at the end of the day I’m not drained, and I’m still able to deliver expeditiously on my obligations.
Before each day, I set up a list of what is expected of me in terms of deliverables from work and projects I am working on. I have also learned to delegate. This is very important when you are working in communities. This allows me to minimize the number of hours I have to spend to be really productive.
I’m trying to prepare for what is coming. I want to fully commit to my education, deliver for my company, and continue working with our communities while still being able to come up for air.
What’s the AI scene like in Nairobi?
In the eastern African region, which includes Nairobi, we are a bit behind in terms of implementation. We are just beginning to understand that we are in a data-driven economy. But at the same time, it’s quite promising. We’re seeing initiatives from our professional communities, startups, research groups, and also a bit of support from the government.
Do you have any advice for people hoping to break into the field?
Before you think about moving into the field, find a domain that excites you. AI is a tool like any other, so you need to know how you aspire to apply it. To take things to the next level, you need to attend conferences and find mentors. Then you need to bring up your technical competencies. Cultivate a habit of reading research papers, case studies, and articles to deepen your craft. I would suggest laying a strong foundation around the mathematics of machine learning, because this is a dynamic field that changes day in and day out. When we talk about machine learning, we have things like multivariate calculus and linear algebra. If you move into data science, you will need to take more classes in statistics. Don’t feel intimidated. A beginner level is okay. As you move up in the field, you will be able to continue improving your mathematics. That foundation is really crucial in this field.
What problems do you think AI is well positioned to solve?
I believe AI can tackle and solve problems in the fields of energy, transportation, food and water security, health care, and alleviating human suffering. We need to ask ourselves what we can do to focus on the United Nations sustainable development goals. Many of us have been trying to emphasize climate change.
How do you keep up on news and research?
I love reading and keep myself updated with the latest AI trends and research. I read a lot of AI blogs, newsletters, and articles. I subscribe to The Batch, which is always on top of the latest AI developments. I also follow a couple of key data professionals on social media. LinkedIn and Twitter have discussions and sessions that help. I follow top-tier machine learning conferences and discussions to gather news about the state of the art. I also follow a couple of threads on Reddit and Quora. I try to look at what kinds of questions are trending.