During my third year of study, I decided to change career path from the field of civil engineering to data science. I applied to many places in Europe, and got an offer to join Amazon for a data science internship in Luxembourg.

It was a valuable internship, I got to play with the huge amount of data that Amazon has, and I worked with supply chain teams on leveraging said data to develop forecasting models that improve the EU logistic operation. I originally planned to stay in Luxembourg only for the four-month internship. But near the end of my time there, I received an offer to become a full time employee, so I scrapped my plan of studying an MSc in Data Science and went straight to the workforce.

After a few years of developing and maintaining forecasting models, my curiosity has shifted from building machine models to building software systems. I reached out to software engineers that I have worked with, and one of their team just so happened to be hiring in Seattle. So I had a chat with the team, got my visa approved, and now I'm doing software development work in America.

On paper, I'm still a data scientist, but I've been doing software engineering work for the past few months and I'm transitioning into a software development engineering (SDE) role.

In the SDE role, my main duty is to build and maintain systems that hosts machine learning (mainly forecasting) models created by our scientists. These models are being used to plan out the Amazon's North America and Europe logistic operations, ensuring our customers get what they ordered on time.

Typically, my work involves designing systems that can address scientists' and other stakeholders' requirements, coding said system, and making sure that it is robust enough to run unsupervised. Also, sometimes when I see an existing problem, I mock up a prototype and share my finding with the team.

In one of my side projects, I developed a "fullstack" anomaly detection system (detection models, backend and a scrappy frontend), which managed to perform pretty well, and it is currently being used to automate some of our data and forecast quality checks. I also enjoy the fast-paced culture at Amazon, where conceptualizing a feature to deploying said feature that affect millions of customers can be done in a matter of days.

During my third year of study, I joined the Hyperloop society, where students from different engineering background work together to compete in the Hyperloop pod competition.

During my time there, I met like-minded people and was exposed to other fields of engineering – in particular, the field of optimization and machine learning (ML) piqued my interest. Then, I started learning programming and computational statistics in my spare time, attending talks given by applied ML and stat practitioners in the University, and managed to start a final year thesis topic that let me apply my newly acquired data science skill set.

From then on, I was lucky enough to get a data science internship in a team that works on a high impact system and promotes the continuous learning culture. Also, through working with other software engineers and getting direct mentorship from senior scientists, I developed my skill set further to get to where I am now.

I also took part in HYPED, the University's Hyperloop student society.

Even though I'm not a civil engineer, the maths and stats skill that I learnt are relevant to me till this day.

Well, I don't have any external recognition, but I'm quite happy that my career switch worked out well and I get to work in a field that I enjoy. Also, personally, it feels good to be part of a team that continuously improves the shipping experience for millions of our customers world-wide. (But don't ask me why when your order is delayed.)

Most people don't know who they want to be when they applied for uni, don't be afraid to change career in the middle of your degree and follow your curiosity. Never stop learning!