Transitioning from Software Engineering to Machine Learning Engineering: Semih’s Journey
Some people are forced into their careers, some choose it outright, but those of us who are more indecisive often end up stumbling into our careers over time. This was the journey for Semih Yagcioglu, the director of Artificial Intelligence at Apziva, and a mentor for Springboard.
Ever wonder what a software engineer really does? Check out Springboard’s comprehensive guide to software engineering. We’ll teach you everything you need to know about becoming a software engineer, from what to study to essential skills, salary guide, and more!
Semih came from not-so-humble beginnings in the Computer Engineering department at Eskisehir Osmangazi University. While he saw the value in computer programming, Semih never felt a fiery passion for the field – that is, until Semih began his senior project…
Semih worked on a unique computer vision program: one that sought to use robots to accomplish tasks that are usually accomplished by human vision like extracting meaning from a single image. To paint a picture, you can imagine what it would be like to go to the MoMa knowing detailed information about every photo or sculpture – everyone would think of you as an art connoisseur. That’s just one example of something computer vision might do.
With a degree in computer engineering tucked behind his belt, it wasn’t hard for Semih to find a job as a software engineer. While the work was informative and certainly paid the bills, it oftentimes felt very robotic (yes, pun intended). Semih longed for a job that would offer the same excitement and intrigue that his senior project once offered.
Unfortunately, back then, Machine Learning (which falls under the domain of data science) was not nearly as popular as the industry has shown today. It was a sort of weird, abstract science nerd among an entire field of science people that nobody really knew much about – yet.
In contrast to programming, Machine Learning works by making inferences and assumptions based on patterns of data to learn how to perform a specific task. So in a field of dedicated computer programmers, the idea of not having to program computers seemed very foreign.
But for Semih, Machine Learning offered a sense of excitement and adventure into a brand new field.
Back to School
Feeling deeply unfulfilled in his work in software engineering, Semih did what anyone bored with work would do: he went back to school. During his Ph.D. program in computer science at Hacettepe University, Semih had the opportunity to work under the supervision of Machine Learning professors on various projects that ranged from natural language processing to computer vision applications. Essentially, they gave computers eyes and ears and said: “let’s see what they can do.”
During this time, Deep Learning, an overarching concept that involves the different fields of Machine Learning began to take off. Just like the weird kid who graduates high school and has a sudden, unexpected glow-up, Deep Learning was quickly becoming the status quo and Semih was keen to join the movement.
Get To Know Other Data Science Students
Learning to Learn
Since Semih began his career in software development before he jumped ship to machine learning, there was a bit of a learning curve he had to overcome. For Semih, he had the advantage of having prior exposure to both software and ML, so in that regard, the transition was not a blind leap but rather a calculated risk, but it still involved some change.
He says, “I think the most challenging part is that you need to get used to designing and training a model to solve your problem instead of coding every detail and case.” Instead of having control over every aspect, you need to trust in the machine’s ability to learn for it to…well, learn.
So although there are similarities between the two fields, it is not always a seamless transition: the tools, terms, and concepts are completely different. According to Semih, “I think in some ways it is a completely new world and in other ways, it is very similar to software development.”
You can compare it to the difference between American and European English; there are different terms, expressions, and meanings in each culture that will never translate directly.
A New Career
Today, as the Director of Artificial Intelligence at Apziva, Semih’s job focuses on finding AI-based solutions to real-world problems and providing consulting on AI to business partners. In contrast to the tedious and predictable routine of coding, the applied research involved in Machine Learning uses a much more agile and flexible approach – one that requires building products around the research that has been done.
Science often requires experimentation to disprove research, but Machine learning revolves around quickly building products and services around the research. It is anything but tedious and predictable – which is exactly why Semih loves it.
Helping other Developers Transition
Through Springboard, Semih has had the opportunity to mentor others in the field of Machine Learning. Like his career work, “it (mentoring) is very very fulfilling in the sense that you are making a huge impact on someone’s life by both providing your expertise in the field but also sharing your experience with them while providing guidance throughout the program.” The goal of this article is to encourage readers who are thinking about changing their career track and desire to become data scientists or machine learning engineers.
Since you’re here…Are you a future data scientist? Investigate with our free step-by-step guide to getting started in the industry. When you’re ready to build a CV that will make hiring managers melt, join our 4-week Data Science Prep Course or our Data Science Bootcamp—you’ll get a job in data science or we’ll refund your tuition.