Q: Julia, you were enrolled in the first year of MIDS. That must have been exciting but also a bit scary since you didn’t know what to expect. Why did you choose Duke MIDS?
A: Yes, it’s been very exciting! I chose MIDS for three reasons. First of all, it’s a two-year program. I thought then and still convinced now that one-year programs just don’t give enough time for students like me with no previous experience in data science to solidify statistical background, Python or R skills, and build a sound portfolio to prepare students for job market needs. I liked MIDS structure that gives a perfect opportunity to focus on the core of data science for the first year to strengthen the fundamental skills. The second year is your time to explore electives and decide in what domain area you’d like to concentrate on in the future.
Secondly, MIDS is an interdisciplinary program. It’s not only offering core data science courses, but also gives an option to choose any elective classes in the areas you most interested in. Starting buidling expertise and portfolio in those domains during the graduate school is a major advantage when applying for jobs.
Thirdly, all of MIDS students are required to complete an internship after the first year. It’s a great opportunity to practice interviewing skills before after-graduation job hunt and to add some real-world projects to your portfolio. I know some students who returned to their internship places full-time after graduation, so overall internship is a major milestone to start your career in data science.
Q: You graduated college in 2015 and took a few years to work. What experiences did you have during those years and what prompted you to go back for a graduate degree, particularly in data science? Would you encourage students to go straight to grad school after undergrad or to take some time to learn about their passions and interests?
A: I worked as a data analyst in the retail industry after completing my bachelor’s degree. However, I mostly worked in Excel doing basic exploratory analysis. Of course, there were some much more to learn in data analytics and data science – that’s why I chose to apply for a graduate degree.
I think a choice to work for a couple of years or to go straight for Master’s degree very much depends on each one’s interests and career goals. I would suggest that a good time going for Masters is when you are absolutely confident in your high-level career interests, e.g., data science or data analytics even though can’t precisely pinpoint if you want to focus on social science or healthcare or business just yet. In this case, Master’s will help you to explore the scape of available opportunities and domains, hone the required skills, and prepare you for the job market.
Q: You completed your internship at RTI during the summer. Can you tell us a bit more about securing an internship as an international student? What advice do you have for other students? (how to search for an internship, how to get the most experience out of an internship, etc.)
A: I started MIDS with a plan to focus on helping non-profit organizations to make a world to be a better place using data science. I later switched to the tech world because of visa reasons (unfortunately, most of non-profit organizations don’t have resources to sponsor visas for international students, but this is a different story).
With a thought to go for non-profits, I already knew that RTI is one of the biggest organizations in the field and had a major interest in joining their team. Luckily for me, MIDS developed a close partnership with RTI and invited two data scientists for Data Talks from RTI’s Center for Data Science with whom I later connected offline to learn more about the jobs. Then, MIDS also had a career session inviting data scientists from the industry to practice mock interviews with us. One of the invited interviewers was an actual data science recruiter from RTI to whom I sincerely pitched my interests about their projects. He was hiring summer interns to work on text analysis problems and I already had a relevant project in my portfolio after the Text Analysis class. This is how I secured my internship! All thanks to MIDS!
To give some advice about how to find internship: first of all, don’t worry! all of you will get one! I think networking is an essential component. Reach out to alumni, connect with Data Talks speakers, cold email recruiters. Cold emailing works surprisingly well: first, go to LinkedIn to find what companies hiring interns at the moment, then find what recruiters might be interested in talking to you – odds are it’s someone holding a title “data science recruiter” or “technical recruiter”. Guessing an email is also easy – it’s probably firstname.lastname@company.com or something similar. Adding a paragraph of why you’re interested in the company, attaching your resume and portfolio with some relevant projects (if any, it’s a plus but not required) will add many points to your advantage and will help you to stand out from people who only apply online. It’s time a consuming process, but it works! This is how I connected with a recruiter from Postmates and got converted to Uber’s employee after Postmates’ acquisition by Uber.
Q: The Capstone Project is the first chance MIDS students get to work with partners on real data problems and essentially get a feel for what it’s like to work in the data science field. You collaborated with the Durham County Criminal Justice Center where you analyzed data across Durham County prisons and courts to investigate racial bias in the Criminal Justice system. Were you always interested in social good/community projects?
A: I started the MIDS program interested in the social projects, but as I mentioned above, had to choose a different path, because non-profit companies often can’t provide visa support for international students. Nonetheless, it was a great experience and I’m grateful to MIDS for matching me with the partners working in the domain of my interests at that time.
Q: You currently work as a data scientist at Uber. Can you talk a little about the projects you are working on? What programs do you use most? Are you autonomous or work in a team? Are you working remotely? What are the pros and cons of onsite and remote work when it comes to data science?
A: I’m working on developing incentive programs for Uber drivers and couriers. I’m analyzing experiments conducted by marketing managers and the CRM team to understand what campaigns would work the best to boost earners engagement on the planform. My work includes experiment design, statistical analysis, causal inference that are mostly done using SQL and Python.
I work in a team of product managers, marketing managers, Ops, CRM specialists, and applied scientists. My role is focused on understanding how to setup the most accurate experiments in all aspects of it – who’s a targeted audience, how to define treatment and control groups, for how long to run an experiment to be confident in the results, expected lift and statistical significance so we could translate it to business needs and adjust the strategy. Once experiment is completed, I would conduct a detailed analysis and share the results and recommendations about what steps my stakeholder should take next to strengthen our incentives programs.
The work is fully remote – I actually met my team only once during the offsite event! During the onboarding stage, I saw remote work as a disadvantage because of so many job nuances I had to learn remotely. Onboarding is a bit harder to do online compared to being physically in one place with my team. Now, however, when I’m getting more confident in the job, I prefer the remote setup. Morning hours are mostly busy with the meetings with stakeholders and the team, but afternoon hours are all mine to dive into projects with no distractions.
Q: What courses prepared you for your data science career? Are there courses/electives you wish you had taken?
A: All of the MIDS core courses helped to build strong fundamentals. As someone working with causal inference and statistical analysis, I think Modeling and Representation of Data and Practicing Data Science are the courses skills from which I use the most at work.
Q: If you had the opportunity to do the MIDS program over again, what would you do differently (if anything)?
A: I wouldn’t take that many electives. I had a fear of missing out if I don’t take as many electives as possible, but it only resulted in lots of stress and medium quality homework sufficient to meet class requirements but not to fully dive into each subject. Less is more in this case!
I also would be choosing electives more carefully. Nearly half of my class signed for the Deep Learning elective, and again, out of FOMO, I took the class too. While the class was very informative, I quickly understood that I have no interest in going so deep into the mathematical theory. I don’t discourage anyone from learning the theory, of course, – it’s crucial, but there are different levels of it. If you are inspired by working on the cutting edge models – this class, and other theory-based courses, would probably be a perfect fit for you; but if you are like me, preferring to focus on the business aspects instead, then choosing more applied courses would be my recommendation.
Q: What were the top 3 best things about the MIDS program and the 3 hardest things about the program?
A: Let’s start with the good part!
I appreciated the opportunity to be together with my class during the entire first year, especially the team-based approach that prepared us for the real-world needs. Now, at work, every project I’m working on is a combined effort of many cross-functional teams, so learning how to work together side-by-side for the entire year was a great experience.
I’m also grateful for the support from professors and faculty members, to whom I reached out for an academical or a professional advice more times than I can count.
Among hardest things, I’d name the workload – it’s a lot of homework, in addition to which you also need to find some time to tune your resume, network, and secure a job.
Q: If you could give any advice to current MIDs students, what would it be?
A: Beyond of what I already mentioned above such as taking fewer electives but more aligned to one’s interests and career goals, I’d suggest spending more time with each other! MIDS is a fantastic place to make friends and meet like-minded people. I wish everyone to spend some time away from the laptops with their new friends, traveling together, exploring Durham, or just having a fun not-data-science-related chat!