Dr. Albert Kahira Njoroge: How it started, how it's going (part 2A)

How did you get into computer science?

My story with computer science is quite funny. I usually remark, I didn't choose computer science, computer science chose me. After KCPE, I selected computer science as my preferred option. Students had an option to select three other courses, that served as backup in case the first choice was not available. I went against the grain and selected computer science thrice, only varying the university. However, at that time computer science was neither well known nor accepted. I was subjected to fearmongering, being told that if I pursued it, my career options thereafter would be confined to working in a cyber cafe. Based on those fears, I chose quantity surveying as my first choice, in the next round of course revision. I later regretted that decision a lot, not because I hated quantity surveying, but because for the first time, people convinced me to go against my gut. Anyway, selections came, and I was placed in quantity surveying.

Back then, the waiting period for university commencement was two years. During that waiting time, my former high school -Light Academy- offered me a live-in position as a computer teaching assistant in the computer lab. One day, my mathematics teacher came to my room and said, “Hey Albert, we found you a scholarship to go to Turkey. You are going to study Computer Engineering”. I had neither applied to the university nor picked the particular course. They picked both the course and university for me. That is how I found myself in Turkey studying computer engineering. That was one of the best things to have happened in my life.?

In simple and elaborate terms, please describe what you do?

My job is an intersection of artificial intelligence (AI), distributed computing and computer architecture. Simply stated, I work closely with domain scientists such as physicists, biologists, chemists and AI scientists to make maximum and efficient utilization of a given computer system. Currently, my work involves working with AI scientists in the pharmaceutical domain to scale and deploy AI models.?

In general, what does a typical day of work look like?

I am not a morning person, never been. I wake up at about 8 a.m., proceed to shower and then treat myself to a nice cup of coffee before stepping out to work. I actually love working from the office but once in a while I work from home. I can't get much done in the mornings, so I use this period to respond to emails and read a few things. I get most of my core work done in the afternoons, though this has been a balancing act of late with increased meetings that tend to be scheduled in the afternoon, so as to accommodate my teammates in the United States.

After a day's work, I head home, rest a bit and then proceed to dedicate about two hours of work predominantly on personal projects such as Datawise and Sisi Academy.?

What does a day-to-day look like in your position as an engineer?

On a typical good day, I write a lot of code. Previously, most of the code I wrote was in relation to training and developing new models. My current position as an engineer involves writing code to deploy models, create interfaces and to define infrastructure (Docker, Kubernetes). I will also do a lot of code reviews, mostly from fellow engineers or scientists. Every piece of code pushed to production is carefully checked by multiple engineers. As this is a highly sensitive domain, we have a lot of processes in place to ensure high quality code and models.?

What about your role as a scientist?

Previously, I held a scientific role, so my work also involved writing and reviewing scientific publications. I still publish, just a lot less.?What often goes unnoticed, when reading a well-written scientific paper, is the amount of effort and time invested. Equally indiscernible is the writing process. Sometimes you write just one paragraph in a day.? I recently just published a paper that I started working on during my master's degree. Since scientific accountability and integrity is mostly upheld through peer review, I also review many papers. I enjoy reviewing because I get to read papers long before they are publicly available. I also feel that it is the best way to keep up with research developments.? I am still hoping to publish or assist in the publication of at least one paper annually.?

Then there is student and intern supervision, one of the most rewarding jobs. I try to take in two interns every year, preferably graduate students looking to transition into research. This can be challenging sometimes, but the work generally involves meeting the students every day or two to check their progress and help solve any blockers they might have.? Currently, I am not supervising any students, but I hope to take up some by the end of the year. My previous company had an internship program where experienced researchers took up interns and the company would cater for their stipends. Such programs are pretty common in Europe and in the last two years I have been trying to replicate the same in Kenya with Datawise Africa.?

What aspects of your work do you enjoy?

My job affords me a lot of learning opportunities. I work with domain scientists whose expertise I know little about. For me to do my job right, they must explain different aspects of their work in detail. This means I get to learn a lot. One moment I am working with a mathematician to understand linear algebra so as to speed up some linear algebra codes in a new system, another moment, a physicist is educating me on the intricacies of dark matter so that they can create an AI model to predict some of its elements. These days, I am increasingly collaborating with biologists and medical specialists, in the process learning about antibodies, drug design and development, clinical trials and a bunch of other pharma-related stuff.?

Podcast recommendation

Methali

51. Kidole kimoja hakivunji chawa.One finger canot kill a louse