‘I just took the plunge and left my pharmacy job.’: Aisha Kala on how she became a self-taught data scientist
AI for Good engineer Aisha Kala talks about her journey of switching to a career in tech and learning statistics, coding and computer science online. A love letter to data and people in AI communities trying to make the world a better place.
I sat down with Aisha remotely for what started as a chat about her decision to change careers, has become an inspiring story of grit and commitment for everyone who wants to work in AI.
In this interview you’ll find out:
- How Aisha found her true calling
- About the rocky road to become a self-taught data scientist
- That the sense of community and the AI support network were essential to finding the strength to pursue her education in tech
- How far you progress once you start applying acquired skills in real life projects
- About Aisha building up a meaningful career in data science
I’ve heard you’ve always had the heart of an engineer, is that true?
I enrolled for Aeronautical Engineering in South Africa in 2006, which had always been my dream. But my father is very old-fashioned and he was dead set against it. He wouldn’t allow it. So I forged my parents’ permission because I was underage. I was quite a rebellious child at the time. So yeah, I started engineering at 17.
It’s quite a tough world out there for a 17 year old to be on your own, without your parents’ support. After a couple of months I went back and I was ready to do what they wanted me to do, which was pharmacy. I can’t say I enjoyed studying it but it got better when I started working. Thanks to pharmacy I realised there’s an altruistic side to me. I like helping people, so that was good. But I needed something more. So I enrolled to study my masters in clinical pharmacy. What I really enjoyed during my masters was my thesis project, it involved collecting data for six months and analysing it.
Was that when you realised you were passionate about data science?
That’s when I was introduced to statistics. Not much of data science yet, because there wasn’t any real modeling involved, but I realised that I enjoyed finding stuff in the data. You know, it’s like looking for that needle in a haystack or going through so much to find this gem. Almost like you’re polishing up the story that was hidden somewhere. That was the first inkling I had that I liked data science.
When did you decide to make a U-turn on your career?
I moved with my husband to Dubai and I started to work there as a pharmacist. I was surprised how advanced South Africa was in clinical pharmacy compared to the United Arab Emirates. The pharmacies in UAE were very advanced in terms of automation, having all these robotic dispensers and electronic medical records, but aside from that they didn’t do much with the heaps of data they had.
Moreover, if I stayed in the pharmacy field, the most I could accomplish was to progress to being a pharmacy director, perhaps a hospital manager or work at a pharmaceutical company. But what I actually like is helping people. And data.
Did you have a plan? Did you know how to do this?
I looked for data science and statistics courses in Dubai, but I didn’t want to pay for a formal degree all over again. I had just paid off my pharmacy master’s and there was no way I was going to do another three or four year degree. I started looking at data science online and there was so much information online, so many videos, so many courses. I thought, well, that’s a starting point.
I started looking at stuff like Coursera and free courses at MIT but soon found that I wasn’t making much progress. It was Covid pandemic and we were working 70 hours a week. It was intense. I would start a course that required 10 hours a week; and I’d give it one hour. By the time I got back to it, I’d forget what I’d done in the previous lesson. It was very frustrating.
So you quit your job.
Yes, I made the decision to stop working. I wrestled with it for a while. One week I’d be telling myself ‘my career is not that bad’; and the next week I’d be like ‘I’m still young there’s still a chance to move to something else’. I just took the plunge, left my stable pharmacy job and I started learning full-time at home.
Wow, that sounds exciting and scary at the same time. How was the jump straight into the tech world?
What I didn’t realise was how little I knew in this new field I’ve chosen. I could do basic stuff with Excel and PowerPoint, but I’ve never learned to code. I started an introduction to data science on Coursera and it kind of threw me in the deep end. I’m learning Python and the next thing, I have this project where I have to create an app or do something by myself, while I just started learning how to code.
My confidence started to dwindle. I had to approach this differently. I downloaded the DataCamp app, where I would play on my phone trying to get points. It makes you feel good, because of all these updates like ‘wow you’ve got 250 points today’ or ‘you’ve done so much this week’ and I did start learning better that way. Python became interesting to me again, after the frustration of jumping into it from the deep end. Once I started learning the basics I realised it wasn’t that bad at all.
“One of the early highlights was creating my own website in a Harvard online course. I never knew I could learn HTML, CSS and a wee bit of Java, enough to create a basic website. I was so proud of myself, sending the link to my mom, to everyone. I can’t remember experiencing such joy as a pharmacist.”
Many people would find it difficult to overcome this kind of frustration on their own. Did you get support from anyone?
At that time, I was looking to network with other people, trying to meet like-minded people; see if anyone else tried a career change. I stumbled across the Netherlands group of Women in AI exchanging over Whatsapp and Slack. I joined the community and it was wonderful. Most of these women weren’t transitioning careers as radically as I was, but there were a few that studied history or majored in literature and then decided to get into AI or become a front-end developer. That was so encouraging. I started speaking to some of these women and it gave me hope that it can be done. I may be slower than others, but I could learn this.
That’s when I met Andreea Moga, the AI and Ed tech expert, during a sort of ‘metaverse’ event I got recommended by the group. I was looking for someone to mentor me at that time. Andrea and I had a connection and she agreed to mentor me. I started a learning program of online courses drawn up by her. She gave me a lot more structure and started to find the gaps in my knowledge. It was Andreea who suggested that I was ready to put my limited skills to use and introduced me to FruitPunch AI.
AI for Good Challenge time! How was it, applying your knowledge in a real-life project?
We’d always speak about my intentions to help people and join a company or community that is doing good. She said: ‘hey, there’s a Wildlife Challenge happening. I know you’re passionate about animals. This is about rhinos from your country!’
So I applied for the AI for Wildlife 2 Challenge. I was terrified when the Challenge started. Everyone I met in the group that I was in was so knowledgeable and they knew what they were doing. But I didn’t let it get me down. If we were given something to do, I would spend up to 12 hours a week trying to figure out what it was before I asked for any help.
I would go to a teammate and say: ‘Look, this is what I looked up and what I tried. I don’t know where I’m going wrong. Can you have a look at my code?” I learned that even though they all had computer science degrees, they were also learning. Sometimes we would look at the code together and they would be like: ‘we don’t actually know where it’s going wrong or how to debug it.’ It’s trial and error.
“I learned that you don’t have to know the answer to everything. Just have the willingness to come forward and do the work. That was half of the Challenge.”
I learned so much from that and I gained a lot of confidence. Any time a term or something unknown would come up; I had this book and I would write everything down from our meetings. I’d look it up and then I’d learn a little more so that I could contribute to my team.
Was the Challenge experience useful? Did it help progress in your studies?
In many ways. After that I actually successfully completed the CS50: Introduction to Computer Science course. It was very, very challenging at times, but it’s also a community of students. It made me realise that I’m not alone and that motivated me to continue.
I also did some statistics and maths courses for data scientists on Coursera, I joined HackerRank and tried Kaggle, too. I wasn’t very successful with Kaggle challenges though. I couldn’t do it in time or I couldn’t complete it on my own. But it didn’t deter me. I knew there would be a day when I can get through a challenge on my own and that day is coming. I just didn’t have the skills at the time.
I’ve seen you showcase final results in the Final Presentations of the AI for Wildlife 3 Challenge. It was for the CI/CD team, right?
When the follow-up AI for Wildlife 3 Challenge came about I was really keen to join. I joined the CI/CD team. It was more advanced than the first one, we were creating a web interface trying to link the containers to the Jetson Nano on the drone and I found that I was less able to contribute in terms of my tech skills.
But in this Challenge, what my team really, really needed was coordination. Nobody had a clear idea of what to do. I thought that I didn’t have the tech know-how to be in charge, but I could direct everything else. I could do the minutes. I could help us stay on track in the meetings. I could keep track of our progress and give reminders and pick up the slack. I really enjoyed the project management part. I wouldn’t mind doing something like that in the future. I didn’t realise I had the skills and now I know it’s something I’m good at!
Ready to put that to work?
During the Challenge I actually started an internship at an e-commerce tech startup here, in Dubai. I’ve been doing some data analysis using power BI and running SQL queries, but I don’t mind it because I’m learning from it. I’m hoping it will be a stepping stone to something closer to data science. And I have enough free time to learn by myself. Right now I am almost finished with CS50: Intro to Programming.
How long has it been since you started the journey of becoming a data scientist? And where do you see yourself next?
In September it will be two years since I quit my pharmacy job. I do think about how far I’ve come. And the whole world, too. There’s is no point in getting back to a job where 80% of the role will be taken up by computers. I’ll stick with this. I’ll become really good in data science and then, maybe, take that to the pharmacy field and make a real difference.
Right now I’m looking for something with a really good team of people that are patient with me learning. I’m looking for a company or an organisation that does some sort of good. Something where I can make a difference, whether it’s people, animals or the planet.
Last question. Which coding language is your favourite?
Some are easier than others. To be honest, I really struggled with C+. I would make one tiny error, a missing apostrophe or a comma and for hours, I could not figure out where I went wrong. That used to be so frustrating. I could break a pencil when I was really angry at this point. I just took a break and came back to it; and within seconds I could see where I’d been wrong. It’s just about changing your perspective. But yeah, I still don’t like coding in C+ very much. I prefer Python. I found SQL quite easy and I don’t even know why, because it’s also a bit finicky. It is funny how we all have different affinity for different coding languages.
Based on an interview with Aisha Kala on 7 September 2022.
