New beginnings, new challenges: my transition from student to full-time employee

Klara Marijan
Photomath Engineering
6 min readNov 21, 2022
Flow of the learning process

Black box. That’s what Artificial Intelligence was to me in high school when I made my first presentation about it. I had no clue about how it works in the background. All I knew was that it intrigued me more than any other field I had encountered. With the knowledge I had then, I only scratched the surface of Deep Learning. I was fascinated by the concept of Computer Vision and the power of giving machines a pair of eyes and a brain to interpret what the eyes are seeing. During the first few years of my college education, I gave it a go with developing Android and iOS mobile applications, tried out frontend and backend development, and learned all kinds of programming languages. As much as I enjoyed expanding my knowledge in different areas, that before-mentioned black box called Computer Vision occupied my mind the most. I slowly started digging deeper than just the surface. I gradually started understanding the mathematics behind it in more detail. The more I learned, the more complicated it got, but the more I wanted to dedicate myself to it. That black box began to unravel itself.

A black box slowly unraveling itself
An actual representation of AI in my mind

And then, I got a phone call which showed me dreams do come true:

“Hi Klara, I am calling with great news! We are very excited to tell you that we chose you for the AI Engineer Intern position at Photomath!”

After a few days of pure enthusiasm about getting the opportunity to work in the Computer Vision industry for the first time in my life, and no less on a product I believe brings great value, I started wondering — what’s my job going to look like? A black box. That’s what it was to me then. Soon enough, though, this one started unraveling itself as well, and what I found inside was more than I could hope for.

I started my half-time job as a student in Photomath in February 2022., just as the final semester of college was approaching. However, this also meant writing my master’s thesis. In spite of knowing I’m going to have a lot on my plate, I nevertheless wanted to broaden my skills even more and pursue my entrepreneurship ambition — by enrolling in FER-s startup incubator — SPOCK, in a team with two of my close friends.

Since the startup we founded was based on Computer Vision, we got first-hand experience in building a complex model that is required to work in real time. Working in a very small team forced us to get to the bottom of every single step in the process — from data collection and annotation to statistical analysis and data preprocessing. We were reading a bunch of papers and implementing the ideas we found in them in order to find the one that suited our problem the best. Finally, we tackled the issue of exporting and deploying the model. During the program, I got a clear picture of how a startup works. Being aware of the difficulties of starting a business made me appreciate Photomath more. By constantly pivoting and redefining the problem and its solution, I got used to frequent changes and adjustments and a certain level of chaos considered normal in early-stage startups. Photomath managed to overcome all that chaos, become stable, and continue to grow.

Photomath team at the company-wide retreat in Poreč
Photomath team at the company-wide retreat in Poreč 2022.

The knowledge I acquired helped me better understand the Deep Learning approaches that we use in the Photomath’s AI team. When I started the job, one of the main worries I had was that I would be doing everything except AI. Of course, some of my first tasks after onboarding belonged to the Software Engineering area. Still, most of what I did was deeply connected to the core of Computer Vision. I had the opportunity to research state-of-the-art methods and implement all kinds of tweaks that might improve the performance of our models — sounds incredible, right?

However, this was when the Dunning-Kruger Effect started kicking in.

The Dunning-Kruger Effect

Somewhere in the middle of the final semester, I was buried deep in the Valley of Despair. There were multiple times I got stuck, and the answers I got from my questions only led to more questions. Luckily, I was surrounded by a patient and understanding team, and they made sure to explain everything I asked to me. I can confidently claim that I’m walking up the Slope of Enlightenment, but honestly, I wouldn’t call it a slope. For me, it’s more of a spiky, oscillating function, but what’s important is that, on average, it goes up.

An oscillating metronome
Me going from “wow, I’m a genius” to “why is my code not working” every 5 minutes

Along with that, I also had to face the challenges of balancing responsibilities and managing time. As the semester was ending, the pressure of finishing my master’s thesis increased, and my to-do list was piling up. Not only was my team incredibly considerate and gave me as much time as I needed, but they also gave me plenty of professional and technical advice. Thanks to their support and my hard work, everything went by smoothly.

I was excited for getting my diploma, and it only grew bigger when I got the offer to continue working full-time. I was no longer an AI Engineer Intern but a Junior AI Engineer. With my new role, I felt anxious about having to meet a higher level of expectations. I wanted to give my best, prove myself and exceed those expectations. In this new world that my first full-time job represented, I held onto the only thing I knew for sure — I needed to work hard. In my free time, I dove into exploring different tools and technologies intertwined with my job and started paying a lot more attention to the details of their implementation and usage. Even now, I’m researching about writing this blog post outside working hours! I didn’t feel pressured or obligated to; I only did it because it interested me, and I enjoy learning new things.

I have only been working full-time for two months now, and I can already see that I have made progress. I have more responsibilities, I’m building trust with other engineers and communicating with different teams. Furthermore, I am gradually working on more complex things.

Of course, there are still times when I get stuck and ask for help, but I believe that it’s all a part of the process. Having excellent mentorship and guidance from more experienced engineers and getting constant feedback about my work makes the transition a lot easier and makes me stay on the right track. There are, and always will be, ups and downs, but staying focused is essential.

So far, I am delighted with my job. I have many new benefits that I didn’t have when I was a student. I feel like I’ve clicked perfectly with my team, and I enjoy spending time with them outside of the office. I am blessed with the opportunity to work on something I love, surrounded by intelligent and creative people from whom I can learn a lot. I think I’ve found the right place for me, and I am looking forward to building my future and growing in Photomath.

Like what you’ve read? Learn more about #LifeAtPhotomath and check out our job postings: https://careers.photomath.com/

--

--