Jieqiu Chen

What do you do?

My responsibilities include developing new data science capabilities, establishing cross-team collaboration to bring the latest technology to our product, sharing best practices, and hiring and coaching junior scientists. Two days in a typical week are fully packed with meetings, and 2–3 days are dedicated to design review, building models, and coding. A few weeks are reserved for learning, attending conferences, and planning. One of my favorite projects was leading a team of project managers, engineers, and scientists from Microsoft’s AI Development Acceleration Program to develop a state-of-the-art system for cross learning on large-scale time series data. I also built a machine learning system to predict the number of users of Microsoft Teams, which has been used to do capacity planning for Teams since April 2020. This ensures we provision enough resources for running Teams meetings to support students, small businesses, and enterprises. What I find most rewarding is that my work contributes to the success of my company and benefits its many users.

What types of skills do you use?

Soft skills: collaboration, communication, writing, influencing. Technical skills: machine learning, optimization, coding (R, Python, SQL).

How are applied mathematics and/or computational science important to what you do?

Applied mathematics is fundamental to my ability to think critically and enables me to abstract a business problem into a mathematical one.

What are the pros and/or cons of your profession/job?

Pro: I find my job fulfilling because my work benefits many people and the decisions made from data analysis could have a multi-million-dollar impact on the business. In addition, there are plenty of challenging projects that I can work on by collaborating with scientists, engineers, product managers, and researchers.

Does your job offer flexibility?

We can work fully remotely or go to the office a couple days a week. Ultimately, we are measured not by when we do our work, but by the impact we bring to the table.

What career path did you take to your current position?

After my Ph.D. I thought I wanted to go into academia, but there were few openings in universities in 2010 after the Great Recession and I was not able to get a faculty job. However, I was fortunate to be offered a postdoc fellowship at Argonne National Laboratory. After completing my postdoc, I realized that doing research was not my passion; instead, I was more interested in solving practical problems. So I went to industry and started out as an operations research scientist at the Dow Chemical Company. While there, I attended a conference where I learned about interesting work being done by “data scientists” (a relatively new job family at that time). I decided that I wanted to become a data scientist, as I found the idea of using big data and mathematical modeling to solve business problems fascinating. I took a few online machine learning courses and started to apply those methods in my daily work. About a year later, I landed a data scientist job at Microsoft and have been working here since then.

Was there anything that surprised you when you started out in your career?

Ph.D. training made me think that solving harder problems leads to more rewards than solving easier ones (rewards = papers accepted to journals, more citations, etc.). In industry, the difficulty of the problem does not necessarily positively correlate with rewards (in the sense of business impact and career growth).

Salary

This is highly dependent on the industry, company, and location of the job. For the tech industry, a good reference is levels. FYI, where it shows the median total compensation for a data scientist is $175K.