Wayfair Data Science Hosts Third PhD Immersion Program

March 15, 2019

Rachel Kirkwood

In January 2019, Wayfair Data Science was thrilled to host our third immersion program for graduate students in quantitative fields. Our largest round yet, this program brought 25 PhD candidates from 12 different universities to spend a week with the data science team at Wayfair’s headquarters in Boston.

Longer than a datathon, but shorter than an internship, Wayfair’s immersion program offers students the chance to dip their toe into the world of industry–without needing to commit to it just yet. For one week, students get a glimpse of what it’s like to work as a data scientist, spending their days much as our team members do: building models to extract insight from Wayfair data, attending deep dive discussions about data science projects in different fields, and getting to know members of the team.

Of course, for those who like what they see, the program also functions as a streamlined interview process for Wayfair’s data science teams. In fact, following the interviews that took place during the week of our January program, our team was so impressed that they extended offers for internships or full-time positions to nearly half of the group!

To learn more about the experiences of the participants, the projects our most recent group tackled, or the structure of the program, read on below! If you are interested in applying for our next round being held May 6-10, 2019, take a look at our application!  


The Program

Wayfair’s Director of Data Science Dan Wulin giving the immersion participants an introduction to the company.

Nearly half of our Data Science team were once graduate students in the same position as our program participants; with this background, our team members have a vested interest in making sure that this program is useful and worth the students’ time. As such, each round we have actively solicited feedback from participants and have used this to fine tune our programming towards helping the participants’ achieve the following goals: improving their data science skills, learning about different possible career paths in data science, and increasing their confidence in entering the workforce. We achieve these objectives through three primary components: the group project, deep dive discussions, and mingling/networking with Wayfair employees and the other participants.


The Group Project

Final group presentation by Emmanuel Ekwedike, Jingkang Zhao, Daniel Sprague, and Liwei Sun.

The core component of Wayfair’s immersion program is the group project. In order to provide enough time to really delve into the data, hone their analysis skills, and get a sense of what it’s like to do this kind of work for a living, much of each day is given over to working directly on a data science project. At the beginning of the week, students are divided into groups, each of which is given a selection of data sets and tasked to use them to solve a particular problem. The first days are spent performing exploratory analysis, after which the team builds and tweaks their models. Finally, they assemble and deliver a presentation in front of their fellow immersion participants and attending Wayfair Data Scientists.

This round, students were tasked to take on one of five projects: predicting expected shipping costs, identifying artists for wall art, detecting fraudulent purchases, predicting revenue and conversion for B2B sales, or using computer vision methods to make predictions about images. Luckily the students didn’t only have data to work with, they were also provided with Wayfair Data Scientist mentors familiar with each topic to offer guidance and assistance throughout the week.  


Wayfair Data Science mentor Tim O’Connor with his immersion group: Supreet Alguri, Alexander Fiksdal, Fadoua Khmaissia, and Harish Guda.


Kirsten Blancato (PhD Student in Astronomy, Columbia University), noted that her “favorite part of the program was definitely the group project.” She elaborated that,

“By the end of the week my team and I were tasked with building a model to predict the probability that an order on the Wayfair website is fraudulent. We developed a ranking system, balancing the order total and the probability of fraud, to prioritize which transactions should be flagged and followed up. Each day my team and I worked collaboratively to try new models and ideas. One of the most useful parts of the week was being mentored by a team of three current Wayfair data scientists. They provided valuable insight into the problem and helped us frame our results in a business context. Even from just this week-long project, I felt like I got a glimpse into what it’s like to be a practicing data scientist.”


Deep Dive Data Science Discussions

Data Science manager Vinny DeGenova doing a deep dive on personalization in data science.

In addition to project work time, a portion of each day is set aside for Project Deep Dives given by members of the Wayfair Data Science team. Each talk focuses on a different subteam (this round featuring Personalization, B2C, Pricing, Marketing, Computer Vision, and Operations Research) and describes in detail one of the projects this team is tackling. These talks seek to illuminate some of the gray area in the infamously broad term “data science,” which is used to refer to large swathes of project types and encompasses many methods. Coming into industry from a wide array of academic disciplines, most students won’t be familiar with all of these variations. Though immersion participants are able to see some of the diversity of data science projects through their team members’ final presentations, those only scratch the surface of what’s possible. These deep dive discussions help to supplement the students’ exposure to the many avenues possible in data science, with the goal that by the end of the week students will not only have a better understanding of the industry, but be able to find their place within it.

This was definitely the case with participant Supreet Alguri (PhD Student in Electrical & Computer Engineering, University of Utah). Supreet commented that prior to attending the immersion program,

“I was actively applying for industry jobs online and was actually very confused where I would fit in industry. So this program was a unique opportunity for me to learn how data science is used in industry and to assess where my background would fit in. The technical deep dives were extremely useful; I got to know how each team uses machine learning to solve a business problem. Thanks to the program I know exactly what I want and which team/area I want to work in.”


Networking and Mingling

Wayfair data scientists and immersion participants mingle during happy hour.

In between the technical presentations and project crunch time, our program also builds in opportunities for networking and mingling. Some opportunities are more formal, such as a panel discussion led by former PhDs about their transition to industry from academia; and some are more informal, like the happy hour with the data science team, the pizza-fueled Thursday hack night, or the side conversations with Wayfair mentors. But all of these opportunities have one goal in mind (besides fun, of course). Our team knows from experience that even if you have the skills to succeed in data science, confidence is key. Getting to know people who have made the transition from academia to industry and hearing their stories can help to build that confidence.

For some, comfort comes in seeing that the data scientist role is not one-size-fits-all. Participant Joseph Reilly (PhD Student in Education, Harvard University) remarked that,

“Talking with current employees about their experiences and how they came to Wayfair was the most rewarding part of the week for me. I don’t have much formal computer science experience and I was worried that I’d feel out of place, but seeing the huge variety of backgrounds and experiences that people brought to the team was reassuring.”

Wayfair data scientists John Walk, Stephanie Sorenson, and Jerry Chen share their experiences transitioning from academia to industry in the week’s PhD Panel, moderated by colleague Licurgo De Almeida.


For others, merely broaching the taboo topic of industry at all is refreshing. As participant Chi Feng (PhD Student in Economics, Carnegie Mellon University) explained,

“Personally as a PhD student, I don’t have much experience with industry. And in my PhD program and probably many others, going into industry is not a topic that’s been widely discussed between faculty and students. However, after communicating with the data scientists at Wayfair who made the transition from academia to industry, and finishing a data challenge with the support of our mentors, I have a clearer idea about what it takes to work in industry and how it is different from working in academia.”


We hope our next round will yield more encouraging stories, as our goal is that every participant walks away from this week feeling more skilled, more informed, and more confident as they explore a career in data science. We look forward to meeting a new group in May!

Apply here to join us!


Submit a Comment

Note: Your email address is required to add a comment but will NOT be published.