It was the first day of the AI apprenticeship program, Batch #2. We had just picked out our seats from the rows of desks available in the AISG office on the 5th floor of the Innovation 4.0 building. I was somewhere in the middle, next to DQ, who was previously a project manager at a large global bank, and Hanifa, who I was learning, as he introduced himself to the rest of the batch, had already been a software engineer for over 10 years. He finished, and it was now my turn to introduce myself.
“Hi, I’m John. I was previously in the finance industry and have been running a startup over the last three years. I picked up data science and machine learning to see if it could help me scale faster. It didn’t work out like I thought, but I found out that I really liked this topic, and so here I am.” I crossed my fingers mentally, hoping that the description of my homebrew attempts at learning and applying AI didn’t sound as insignificant as I felt it to be, surrounded on my left and right by the decades-long combined industry experience of my row mates.
There wasn’t much time to ruminate over it. The AI Engineer mentors distributed the deep-skilling materials right after introduction, and we dove right into the first leg of the apprentice program.
Eight weeks later, I had been ‘certified’ as an Apprentice AI Engineer, and I was assigned to a medical project with another apprentice, Basil – to apply machine learning techniques to electronic health records. Our mission was to predict as far in advance as possible if a patient’s condition was going to worsen. We were given a broad mandate – explore an entire hospital’s worth of data, find insights, and build models to operationalize those insights – which we did over the next seven months. We iterated through several hypotheses and models, and eventually settled on a technique that we had ‘independently rediscovered’ together with some other apprentices on another project (the rolling window + random forest method for time series data). It had the best bang for the buck, working well on small datasets but also achieving a 70% accuracy rate in a three day prediction window. I was very pleased the day that we released the model to the doctor overseeing the project.
But as our batch graduated, I realized that I didn’t really want to be ‘just’ an AI Engineer. I had a deep set of customer facing skills that would be a waste to abandon. So I asked the directors at AISG if there was a role that could combine both my skillsets, to which they said ‘Yes! You can join the industry development team here as a Pre-sales Engineer’. I thought it was a great opportunity to see more AI projects across different industries in Singapore. So, while the rest of my peers went off to launch their careers in AI Engineering, Data Science and Project Management, I spent the next couple of years working with our account managers and industry partners, looking at business challenges, and figuring out how to turn them into AI problem statements. It was exactly what I was looking for, and I got to scope and initiate projects across government, MNCs, SMEs and startups. I saw for myself the abundance and idiosyncrasies of opportunities in AI across all sectors, and as part of the outreach team at AI Singapore, I also clocked a personal milestone – helping more than 10,000 people take their first steps into AI through the AI for Everyone and AI for Industry programs.
Today, I’m at Temus, a Temasek subsidiary focused on enabling digital transformation in both the government and private sector. I’m a Technical Delivery Manager in the Data and AI team – where 70% of us are ex-AISG. It’s a highly technical role, where I’m responsible for coming up with the solution for a client’s business needs, and translating that into POCs, roadmaps, technical architectures and tangible deliverables together with our engineering lead. With the release of LLM technology into the mainstream, both AI and data are becoming an increasing important part of the enterprise tech stack, and every day I get to be a part of AI use cases that are helping to drive industries and our economy forward.
Of course, none of this could have been possible without the launchpad provided by the AIAP program, and every time I connect with AIAP alumni, I see my story repeated in their own paths – as AI Engineers, Data Engineers, Data Scientists and Product/Project managers. I love the impact that I’m able to make with this technology, and if you’re thinking this might be where you want to take your career, I’m hiring!