My journey as an analytics intern with Saatchi & Saatchi Wellness was certainly one I could not have imagined.
As a senior perusing a degree in computer science at Hunter College, I was given a chance to have a series of ten-minute interviews with a variety of companies for a chance to land an internship for the spring semester as part of the TTP Residency @ Hunter College program. The TTP Residency @ Hunter College is a semester-long program that is part of the CUNY 2xTech initiative, designed to double the number of graduates from the CUNY system with tech Bachelor’s degrees by 2022.
Naturally, I had feelings of extreme nervousness as this was my first interview of any kind with the potential for a career-defining opportunity hanging in the balance. Not knowing what to expect from such an abbreviated interview process, over the weeks leading up to my interviews, I did my best to prepare by studying some of the most difficult technical problems that I could have possibly been asked.
Finally, came the day for my round of interviews – and I was shaking. The nerves got to me on my first interview and I felt as if I did not perform as well as I could have. In my second interview, I was interviewing for a position that wasn’t fully aligned to the type of work that I wanted to ultimately do – which led to a few awkward moments. At this point, while it may have felt like nothing was going my way at the time, the reality was that those first interviews were invaluable learning experiences. Based on those learning experiences, I felt much more confident as I prepared for my third and final interview with SSW. Right from the beginning of my interview with Heidi Wu (who I later learned was a former SSW intern who was now a full time employee on the team) and Vlad Ryvkin (one of the senior data scientists on the team), I felt some really good feelings about this team and agency. The technical questions that I received, while a bit challenging, were relatively straightforward compared to what I spent the previous night feverishly studying. We ended the call on a positive note – and the weeks following were full of wishful waiting and hoping. Then the call came: to my relief — I was matched with SSW.
Fast forward a couple of months and it was my very first day working as an analytics and data science intern with SSW. I was greeted by the most amazing team I could imagine and I was working in a beautiful high-rise office in midtown. From the start, I was given a lot of flexibility and the team was very gracious in providing help with anything I needed – my first day couldn’t have gone any better. After getting a grasp on the day-to-day operations within in the team, I was given a project that leverage supervised learning methodologies that really tested both my skills and patience. Throughout the weeks I spent working on this project, I found a new best friend named Google. I didn’t expect to learn so much from being a challenged by a problem that I thought I can get through with ease.
As if my internship wasn’t eventful enough, we were hit with an unforeseen pandemic. COVID-19 shook up the experience of my internship in new ways. I had to learn to adjust to the realities of working from home as an intern. The change in atmosphere was a new challenge to deal with – especially because I wasn’t able to experience the close proximity to the team that I had started getting used to in the first few weeks of my internship.
However, the team challenged me with a new project more fitting with the situation. The task was to run an analysis on COVID-19, and how the pandemic was affecting different parts of the country based on a multitude of Social Determinants of Health data. After researching a multitude of publically available data sets (both related to epidemiological dynamics, and geographically linked social and economic data sets), I decided to focus my project on ways to use the CDC Covid-19 tracking data and CDC’s social vulnerability data in conjunction with supervised learning algorithms (such as Logistic Regression and XGBoost) to better understand what factors were highly correlated with high infection rates, and predict which US counties would have certain proportions of their population be infected with the virus.
The map below is an example of one of these analyses, where I show which counties have 1% of their population infected versus what counties I predicted to be infected at that level.
The potential use for this project would be to help identify which counties are most vulnerable to outbreaks and how can we improve their status. The project gave me the chance to learn a new technology, create a portfolio worthy project, and helped me get comfortable with public speaking after presenting it to my team. If you’d like to learn more about this project, and the methodology that I developed, please check out my article on Medium where I discuss my whole analysis.
My entire experience as an intern with SSW analytics was one that I am very thankful for and one that thought me more than I could dream of. It allowed me to confirm my passion for data analytics and helped me strive to new levels. Thank you to my whole team for giving me priceless guidance, lessons, and showing me what a real team looks like.