Sophy Toteva

Could you tell us a little bit about yourself and about your studies before coming to Imperial?

I'm originally from Bulgaria, and came to England when I was 13 to attend boarding school in Brighton.  Afterwards, I went on to study maths at King’s College London, where I mostly focused on topics within the field of theoretical physics.  I love the subject because of how abstract it can get, so maybe unsurprisingly my favourite module I ever did was Group Theory.  Likewise, I also had the chance to pick up a couple of computing modules, where I first got exposed to programming.  I found the problem-solving required to build an algorithm ignited that same curiosity my beloved Group Theory did.

What attracted you about the MSc in AI?

Two of my all-time favourite subjects to read about are maths and medicine.  I never knew how to combine them before so I often had to pick one over the other, for example when I was choosing what to study for my undergraduate degree. Then, during my final year at King’s, I discovered the Algorithmic Human Development research group at Imperial and found their research rather captivating.  I was amazed by the innovation technology could bring about when mixed with two of my main interests, and thus found the MSc in AI to be the perfect place to combine everything I love.  Moreover, the curriculum seemed to offer a very thorough transition into the field by covering a vast range of very fundamental subjects for becoming a great ML engineer.  Additionally, the MSc in AI is specifically curated to take you from a beginner to a solid programmer in the span of a year which was perfect for someone like me who had very minimal experience with Python.

What did you enjoy the most?

I really enjoyed the second term.  It felt like an amazing milestone during the MSc as it was at the point where you had gotten much more comfortable with the workload but also had made some really great friends.  I will always look back very fondly on the late nights in Huxley with the other students, exchanging ideas and learning from each other.  It felt like I was part of something really special and it was truly the most fun and fulfilling time on the MSc for me.  It was also the time I met my boyfriend, so I guess that was an added bonus. 

What did you find more challenging?

The first term definitely feels quite demanding, especially if you are a beginner in Python.  The coursework starts straight away, but although it was hard to keep up at times, I also felt it was the most effective way to improve my programming skills.  The Python module is very well structured, in my opinion, and in the span of two months you begin to see huge improvements in your abilities! 

Could you tell us about some of your achievements on the MSc that make you proud?

I did my individual project on quantum computing, and this extended the taught part of the course quite considerably.  I am proud of this achievement because for a moment at the beginning of the project I felt like I was back at the start of the year: with little knowledge and much to do.  Nevertheless, with amazing support from the department I was able to grow as a researcher and it inevitably showed me what kind of work I wanted to purse after graduation. 

What did you do in your spare time?

I mostly hung out with other students from the cohort.  The MSc is quite workload-heavy so I didn’t have a ton of spare time, but it was the same for everyone on the course so we shared the ups and downs together.  Often we found ourselves at some pub near college on a Friday or meeting up for brunches or a gathering at someone’s house.  The MSc AI coordinator, Rob Craven, also organised MSc picnic meet-ups with current and past cohorts and we got to play rounders during the heatwave last summer. 

Could you tell us about your individual project?

My individual project was on quantum computing—quantum simulation of the Ising model.  It was an interesting process initially as I kept proposing topics which simply would have needed much longer that 4 months to execute.  However, this pushed me to see the more granular elements of complex research problems—something I am finding very useful in my current job.  Overall, I loved this part of the course as it allowed me to study one of my favourite subjects in the world, namely quantum mechanics, in a much more practical manner—there is genuinely something for everyone on the MSc in AI! 

What have you been doing since you graduated?

Initially, I joined a consultancy as a data and AI engineer.  However, I felt the job consisted of more data than AI engineering and I had really started to miss the work we did on the MSc.  After about two months, I moved to a company that does AI-based compression as a research engineer and have been absolutely loving it.  My day-to-day schedule incorporates the perfect mix of mathematical theory and computer programming and has proven to be the perfect place for me to grow as an ML engineer.  It almost feels like I am back on the MSc—producing cool research with some very clever and fun people! :)

Do you have any advice for prospective students?

When Josiah Wang (the Python Module lecturer) suggests you do the first few Python lessons before term begins—you do them!  I would also suggest to try and stay on top of current ML research by reading papers, due to the fast-changing nature of the field. What is popular today is often outdated by the following year, so in terms of being prepared for industry, it is nice to have a more rounded knowledge of research topics in AI.  Additionally, you may find some of the research is applicable to the coursework, and trying to implement new innovations can always prove to be a fun exercise!