Jamie Fradkin has worked at BuzzFeed as a data scientist for over a year and a half after attending Metis data science bootcamp in New York City. As a former biomedical engineer, she wanted to switch careers so that she could make use of her math and statistics skills in a booming new industry. Learn about the Metis application and learning process, Jamie’s experience getting hired and changing roles at BuzzFeed, and her tips on building confidence in your own skills.
What's your career and education background? How did your path lead to a data science bootcamp?
I went to Johns Hopkins and studied biomedical engineering and applied math and statistics. My first job out of college was at a medical device company where I was working in research & development making surgical tools, doing lots of mechanical work and quality testing. I was really passionate about that industry and felt it was something that would be a great fit for me, but I wasn't making use of the statistical or math foundation that I had built up in college.
I heard about data science being a booming new career that had more demand for my skills and interests in a variety of industries and locations. After about a year working in the medical device industry, I decided to make the change so that I could best leverage my education, and have more opportunities and options for my career.
I felt a bootcamp was a super smart and efficient way to go about leveling up my data science education as opposed to going back to school for a formal master's degree. In my research, I found that Metis had a really comprehensive curriculum with experienced instructors so I thought that would be a good fit for me.
Did you consider going to a bootcamp for any other digital skills? What attracted you to learn data science?
In a lot of technical fields, you can feel a bit out of touch with the actual strategy behind decision making. You provide reports and are the go-to person for technical questions but you aren't really in the loop on how that's affecting business operations. Data science is a great bridge between those hard quantitative skills and soft skills. It allows you to work with a lot of different kinds of people – you’re able to communicate your ideas to a non-technical audience for them to gain the most insights possible from their data. Data science is a really cool bridge between the two worlds – it’s a unique job.
How did you find out about Metis? What stood out about the bootcamp and did you consider any others?
My dad sent me an article about how college graduates with humanities degrees were going to coding bootcamps in order to develop technical skills that would make them more appealing in the job market. Even though I had two varying technical degrees, I didn't know anything about machine learning or visualization. I learned about Metis from my dad and it came at a great time.
I applied to Metis and Galvanize, but it came down to location because I wanted to be in New York. I read Course Report quite a bit before I started and it seemed like a lot of people had phenomenal experiences so I was sold.
How was the Metis application and interview process for you?
Metis changed the admissions process a bit, but when I did it it was really open-ended. We had to craft a data science project and articulate to the interviewer how we would execute it. And of course there was a Python and SQL test. I liked how they tried to determine your curiosity, creativity, and grit, which is Metis' tagline. The process tested what kind of data scientist you would be if admitted. I'm not sure exactly how competitive it was, but I thought the application was pretty fair.
How was your Metis learning experience? Did the teaching style match your learning style?
I went to Metis in January of 2016 and I started my new position at BuzzFeed in May of 2016. Metis is 12 weeks so literally, every day counts. You can't slack off and you can't miss a beat. The structure of every day was perfect for what I had envisioned – you have a morning lecture and then the entire afternoon is time to work on assignments, projects, and ask your teachers questions.
Every morning there was a lot to digest, but you could always ask your peers and instructors for help. It wasn't a typical college setting where you get a lecture and then they hope you understand it by Googling it. I felt there were endless opportunities to make sure we fully understood the topics we covered.
I loved how collaborative the experience was. In a bootcamp, you really bond with your peers. Everyone is having a tough time because everyone comes in with different knowledge. It certainly wasn't competitive during the course because we all wanted to make sure everyone was up to speed. I had a great time.
How did Metis help prepare you for the job search?
I think similar to what I described about the actual curriculum at Metis, the career support unit provided endless opportunities to ask questions. So while there's no system where a career counselor is going to literally match you with a job and get you hired, I felt I had all the resources I could ever need to accomplish that on my own. We could get advice on what to wear to an interview, how to write an email to the hiring manager, or see if the Metis alumni network had any connections in a specific company. The career counselors were your friends. Metis gave a lot of general career education to bootcamp students as a whole, but they also catered to you on an individual basis to make sure you found the best fit for you.
Do you have any advice for bootcampers who are going through the current job search?
The people who have the most success looking for jobs are the ones who have a personal reason as to why they want to work at a certain company. You can spread your resume to as many places as possible and hope something sticks, but when I made it pretty far in the interview process or even got offers, it was really because I genuinely had an interest in the company's mission, and I really felt my skills would be a good fit for them.
It's important to take time to think about the kinds of companies you want to work for and hopefully, your interest and your skills will shine through.
You’ve been with BuzzFeed for about a year and a half. How did you find the position and how was that interview process?
I found it by cold messaging a recruiter on LinkedIn, which I think I had done quite a bit back then. It actually turned out that another Metis graduate was already working at BuzzFeed – she was a phenomenal contributor to the company and had really excelled since starting there. So I think that actually helped my case quite a bit.
One or two years ago, bootcamps weren't really a well-known concept yet. The recruiter connected me to the Metis grad and there was a phone screen. I also had an in-person technical interview, another virtual technical interview, and then there was a take-home assignment, which is where I was provided a data set of lots of stats around how users engage with BuzzFeed quizzes. It was very open-ended where they only said, "tell us some insights about this data." I made a presentation that was designed for a non-technical audience, which might mean people who write quizzes or people who work on distributing it across all social media platforms. My task was to essentially take a massive data set about how users interact with quizzes and give them some strategic recommendations based on that data.
The interview process at BuzzFeed was long and challenging – around six weeks. But like I said before, I passionately loved BuzzFeed so before I started there they were able to see that. The take-home assignment was enjoyable for me.
Describe why BuzzFeed needs a data analyst or a data scientist. Could you tell us about the day-to-day of your role?
I’ve had two separate roles since working at BuzzFeed. One was more on the data analytics side, which is necessary at BuzzFeed because we have a huge team of content creators – people who write posts and make videos and quizzes etc. It's really important for there to be a feedback loop between our audience and our content creators so that we know what's truly working and connecting with our audience.
So in that respect, a data scientist is really helpful to sift through all the page views, shares, comments and other engagement benchmarks to figure out what's working for us as a brand. Another side of BuzzFeed is we have a lot of what we call owned and operated properties like a website and app. I'm now working on the BuzzFeed app as a data scientist where I provide a lot of reporting and metrics on how users are engaging with content to help inform new designs. We are content agnostic because it's not about trying to decide what content to feature in the app since that's all done automatically. In the app we can control the design, the layout, and other features that we want to have. One main area of BuzzFeed data science is content and one area is products such as the site and app. Another is distribution, which is thinking about what we put on Facebook, Twitter, Instagram, Snapchat, etc.
Are you using the same languages and tools that you learned at Metis? Did you have to learn any new technologies to work at BuzzFeed?
Language-wise it's always Python and SQL, and we use Jupyter Notebooks just like at Metis. BuzzFeed does have a couple of in-house business intelligence tools that we're allowed to use. All in all, there's a ton of overlap with Metis. The only thing different at a larger company is that you're exposed to more production-level tools like Spark and Hadoop. When you need something to be fast and efficient for millions of reviews it's got to be production-level machine learning, whereas at Metis it was only small data sets.
What was the learning curve like for you when you first started at BuzzFeed? A year and a half later, how do you feel you've grown as a data scientist?
Part of what happens when you first start at a company is needing to get that domain knowledge. BuzzFeed is part of the media news and entertainment industries, so I had to learn what metrics matter to people and how we define success here. I had to learn what information to provide to our different stakeholders to help them best do their job. I've had a lot of great opportunities to extend my technical skills, take classes, and collaborate with other fellow data scientists.
I think I've grown in the way that I communicate this information to other BuzzFeed employees. I’ve become familiar with the data that we have and how I can best serve everyone’s needs.
Does BuzzFeed do a good job of onboarding their new hires and making sure that their data scientists are continuing to learn and grow?
Around the time I joined BuzzFeed they were really interested in shifting the concept of a data hire from an analyst to a data scientist. We wanted to set that expectation of what kind of levels someone will be delivering at. It's really about providing results based analyses. The data science department has definitely grown quite a bit since I've started. We’ve hired between five or 10 new people in the past year.
Another thing that I love about data science at BuzzFeed is the opportunity to have side projects. There isn't too much restriction on what you're allowed to work on because the leadership here realizes that data science is actually a creative pursuit. So if you have a curiosity about something that wasn't directly assigned to you, there's so much room for experimentation without negative consequences. We all learn from each other and share findings all the time. I think mentorship can come in a lot of different forms and it doesn't necessarily have to be someone who has a higher education level than you or even higher experience. It's just learning from someone who has different skill sets.
In terms of your background in biomedical engineering, has any of that experience been useful in your current position as a data scientist at BuzzFeed?
Knowledge-wise there probably was not much overlap. There is a fearlessness that I associate with that time in my life – not getting knocked down by any seemingly impossible task that gets put in front of you. I learned to have persistence and essentially be able to start a problem without conceptualizing what the solution would be. That's where I'd say the two fields overlap.
What role do you think Metis has played in your success? If you didn’t attend a bootcamp like Metis would you be where you are today?
It's hard to say what would’ve happened had I not gone to Metis, but I think it really helped to have an accredited program on my application. I technically could’ve just googled and looked up all the topics that Metis covered and maybe taught myself, but they had a really phenomenal reputation. Coming from that program certainly helped my case quite a bit.
And for some reason, Metis instilled a lot of confidence in me even after a really short time studying in this field. The curriculum was so perfect because you did homework assignments and projects presentations that were similar to what you would really do in a data science role. I had the confidence to know that I could do this, I just had to find a company where I would love to do data science.
Are you still involved with Metis alumni and instructors?
Yeah. The first year or so after I graduated I went to career days and I was an alumni interviewer. I went to a lot of open house events and things like that. Every time they ask me to participate in something, be a speaker or connect with one of their students, I always say yes because I think it's really important to maintain the alumni network in this community. So I'm still relatively involved. I really feel that I owe them a lot for getting me to the level where I could have a job that I love.
What advice do you have for people thinking about making a career change and attending a data science or coding bootcamp?
I get this question a lot and it's hard because data science and data scientists are still terms that mean so many different things in different settings. My initial advice – don’t choose this career path because you think you should. Choose it because you want to and you think it would be a good fit for yourself. I'm assuming that whoever is going through this process has done enough research and knows that it's a career path that would be suitable for them.
The other component of that is to not be intimidated by data science because there are a lot of scary graphs and terms. But I really feel that in a bootcamp setting, as long as you have the willingness to work hard and to learn, you can succeed with the support that a bootcamp provides. Essentially, if you're considering applying and you think you'll enjoy it, don't be swayed and don't be intimidated – have confidence.