Carnegie Mellon University

Molly Zhu

Molly Zhu

Data Science and ML Engineering Manager - Spotify

When did you first become interested in quantitative finance?  

Quantitative finance is a hot area that can lead to a prosperous career. I had always been interested in mathematics and thought the field could be a good match for me. In high school, I participated in anything math-related, including Math Olympiad, and my best bet was to study something that was quantitative and analytical. I was not determined to pursue a mathematical research career and decided that I had to find another area where I could apply my quantitative skills. Even though I lacked specific knowledge and personal connections in the industry, I thought the financial industry would be interesting and provide me with a solid job, so I went for it.

What led you to Spotify?

Over the years, I became more and more interested in tech. Growing up with my dad as an engineer, I always had an interest in the area and was attracted to innovation. I love building things, and I like consumer tech. After the financial crisis, the financial industry became heavily regulated, and rightfully focused on less derivatives, which, however, is where the quantitative knowledge really applies. I found my work becoming less interesting than my peers who worked in the booming tech industry. As some of my classmates moved in that direction, I thought I should give it a try. The fun, collaborative, and music-rich culture at Spotify basically had me at the hello and I’ve been very happy with my decision since the move.

What are some of your daily roles and responsibilities?

There are several major components to my job, but overall, I build data pipelines, gather a significant amount of data, conduct analysis, and build machine-learning models for Spotify’s ad business. We conduct analysis to make Spotify profitable through the free (ad-supported) business,  optimize  our campaign strategy, and estimate the financial impact of undesirable behaviors on the platform. We build machine-learning models to help brands expand their target audiences based on look-alikes and deep learning models for audio recognitions. We also contribute to Spotify infrastructure and efficiency hacks by building tooling libraries, templates and reusable code.

What do you enjoy most about your career?

I really enjoy playing a role in helping artists live off of their art. I appreciate art and the artists who spend their time and effort creating it. Unfortunately, most artists, including musicians, cannot make a significant amount of money from it. This inspires me to work harder every day to help support those artists. This is part of Spotify’s mission, and it’s a shared value across the company. The work is also very challenging and cutting edge, which I find rewarding.

How has the MSCF program helped you in your career?

I came from a theoretical mathematical background and learned most of my programming and computational skills from MSCF. Prior to MSCF, I had limited programming experience and the program helped me to build my skills. In my current role, I am authoring and contributing to production libraries, which relates to the work I did in financial computing courses at MSCF. Analytical works using simulation and statistics are also part of my job. The financial computing course by Professor Dmitry Kramkov has had a profound impact on me throughout my time in the tech industry. The concept of design in large-scale programming provided guiding principles for my work in this area.

What is your advice to those considering a master’s degree in financial engineering?

Even though I did not stay in finance throughout my career, giving it a try was absolutely the right choice for me. I hope prospective students know for certain where their interests lie, finance or tech – but for anyone who is unsure – a masters in financial engineering is a great thing to try. If finance does not end up being the right fit, the program is still very helpful if you want to go into tech.

What attracted you to MSCF compared to other programs?

The program was ranked number one, had a great reputation for job placement and for being the most engineering-focused. When I received the offer from MSCF, I immediately accepted it. For me, since engineering was my weakest area, the program was the best suited to help me improve those skills and broaden the width and depth of my knowledge. The coursework, career services and locations were also influencing factors in my decision. 

How would you describe the ideal student for the MSCF program?

The ideal student for MSCF must be eager to learn, curious and determined. The program is tough and all students faced challenges throughout their courses, so students need to believe that they can do it and put in the effort in order to do well in the program. Students will also need to have patience and perseverance, as much of the knowledge doesn’t come overnight and takes a lot of practice. They have to be persistent and keep doing what they’re doing. Do all the hard work, and you will be rewarded. 

What was the biggest surprise during your experience in the program?

I was surprised by the degree of focus on engineering and computing and that I had to write a lot code! A lot of students in my class were engineers, and I was probably one of the few students who entered the program without an engineering background. Since I had a strong background in mathematics, theoretic courses were less difficult compared to the computing courses. However, other students often found the courses like stochastic calculus to be much more challenging than programming courses. It took discipline and patience to learn programming well. It’s been incredibly helpful in my career and ultimately allowed me to transition to tech and do what I do now. 

Anything else?

I just think the program is really great. It’s very challenging - more challenging than I expected. When I was in China, the perception of American universities was that everything is very easy – just apply, go and graduate – but that was not the case at all. I was pushed really hard. I was pushed to find my limits, my boundaries. It was really hard work and I really appreciate it.