Unlike many web development bootcamps, data science programs often require some background knowledge in statistics and coding. So we’re diving into the admissions process at data science bootcamp Metis, and getting insights from alumni Deepak and Emily, who explain their journeys from the Metis interview to their new jobs at Facebook and Etsy.
Amy, tell us about your role as the director of admissions at Metis.
Amy: As Director of Admissions, I'm usually the first interaction students have with Metis. I walk students through the different stages of the admissions funnel and work one-on-one to help them determine what course will be a great fit for them. Sometimes it's the immersive bootcamp; other times it's our professional development courses.
In some cases, when they're not yet ready for the immersive bootcamp, I provide resources to help them build up their backgrounds. My goal is to help students create a career path of success and help them move along the application process as seamlessly as possible.
Can you walk us through the steps of the Metis admissions process?
Amy: There are three steps of the application process.
- Submit a written application. There are open-ended questions and Python programming exercises. We ask students to have some programming experience and a statistical thinking background, and we’re able to see that in the first round of the application.
- Coding Challenge. If they move on to the next round, they are given 48 hours to complete a coding challenge. The challenge involves a technical assessment, an exploratory data analysis, and a data science project.
- Interview. Once we have those materials, we’ll set up a 30-minute interview with one of our data scientists to get to know the student more, help us determine if they are a good fit for Metis, and see if we are the right fit for the student.
It takes two to three weeks from the time you submit an application to receiving a decision.
What level of programming or statistics do you want to see in an ideal applicant?
Amy: We've intentionally left this pretty open ended. Although we don't require students to have a background in Python specifically, we want to see how you tackle something you might not be familiar with. We do expect that students come in with some programming experience and some background in statistics, and we've designed the admissions process to delve deeper into those areas. You don’t need to have worked as a web developer, but if you don’t like programming, then this is not the best profession for you. We can test these skills in the admissions process based on the questions that we ask.
What sort of backgrounds and experience in coding do Metis students have?
Amy: Students come to Metis with various backgrounds. Some have the usual computer science, economics, math, and physics backgrounds, but then we also see students with non-traditional paths to data science: finance, history, English, or psychology.
One thing our students have in common is their interest in technical skills and their statistical thinking. You may not have taken a formal statistics class, but this is something that you're very curious about, and you're able to show that your path has lead you towards this career in Data Science, even if your degree doesn't show that.
We have about 60 hours of prework to help students develop a stronger foundation in areas that they might be weaker in. This includes Python, linear algebra, and statistics.
Are presentation skills and communication skills important to being a data scientist?
Amy: Data scientists, in general, need to be strong communicators both visually and verbally, because data scientists are really the translators within their company. They're taking the insights that they find in the data and communicating them to a wide audience of folks who may not be familiar with a lot of the technical aspects but need the insights from that data to make decisions and drive value.
During the bootcamp, we focus on five different projects where students actually present their findings, because we want them to be comfortable communicating their findings in front of people. During the interview process, we have a project challenge that they actually present to our instructor.
Emily, I'm curious what you were up to before you went to Metis?
Emily: I came straight to Metis after getting a Master’s in Organizational Behavior, and before that, I got an undergraduate degree in Decision Sciences. I did have a lot of statistics experience, because I was a Statistics minor in college and I took econometrics courses in grad school. Conducting psychology research also requires a lot of statistics.
I definitely didn't have a traditional computer science background. I had been using R, which is a language many data scientists use. I had taken one class in Python, and that was four years ago.
What I really liked about Metis was that it filled in those gaps I had in Python and machine learning. I hadn’t done those more advanced techniques beyond regression.
Did you do your coding challenge in Python or in R?
Emily: I did do it in Python. I had a little bit of exposure to Python, and it is somewhat similar to R. I definitely needed some online references, but just to look up how to do an IF ELSE statement with Python syntax. I also knew that the Metis bootcamp was going to be in Python, so if I wasn't comfortable completing this coding challenge in Python, it probably wouldn't have been a good fit.
Was the 60 hours of Metis prework helpful to you?
Emily: Definitely. I was strong on statistics and weak on computer science. There were some students with opposite backgrounds. So the prework was a good refresher, but I really needed the Python, computer science, and machine learning focus.
Deepak, how about you? You had actually worked as a web developer before but you more recently working in education. How did you decide to make the switch?
Deepak: My background is in math as an undergrad, and then I started working with databases. I was working on end-to-end projects, analyzing and visualizing data, and presenting that to an audience. During that phase, I had to transition to a web developer on my own, but throughout that time I was working with a lot of data.
Even during grad school in economics, I had a lot of exposure to economic data. So I've been playing with data for a long time, and that's what motivated me to come to Metis. Although I had a lot of data analysis skills, I didn't have a lot of computer science, machine learning, or predictive modeling skills.
Could you get through the Metis take-home challenge with your Python skills?
Deepak: To be honest, I didn’t have any Python skills before I started the take-home challenge. When I spoke with Amy, she said, "You can code the take-home challenge in any language you want, but we recommend Python." So I started looking at Python. It was a steep learning curve for me, but I did my take home challenge in Python. Because I had knowledge in other programming languages, I was learning the syntax.
What did the Metis application tell you about the Metis bootcamp experience?
Deepak: The intensity of the application gives you an idea about how rigorous the bootcamp will be. I had looked into other data science bootcamps as well, and some of them seemed like they would take anyone. At Metis, the take-home challenge and the interview process was challenging. You really had to put in effort to get through it, and when I did get admitted, I felt like I deserved to be in the program.
Emily: I had a very similar feeling. Also, my brother is a data scientist; he'd spoken at Metis before and been to their career day, and had great things to say about Metis. I trusted his judgment. He told me that after 12 weeks, Metis students were giving really good, really strong presentations. That gave me a great feeling about Metis.
What stood out to you about Metis when you were looking for data Science bootcamp? Did you look at other data science bootcamps?
Emily: Looking at the landscape, there aren’t as many data science bootcamps as there are web development bootcamps, so I didn’t have to compare 20 different schools. I knew I wanted to work in New York, so that limited it further. I looked briefly at other ones, but I knew Metis was competitive. I looked at the syllabus, and Metis taught exactly what I wanted to learn: machine learning and Python. I looked at NYC Data Science Academy, but they covered some material that I already knew (they teach R as well as Python).
Deepak: In my research, start date was important, and I looked at where alumni landed jobs; it looked really strong. The key factor for me was talking with the Metis data scientists during the interview. I had a chance to ask them honest questions, and they were really honest and open in their answers.
Did either of you want to work in a specific role or at a specific company when you graduated from Metis? What was your goal?
Deepak: I was not looking at a specific role or a specific company. My goal was to gather strong data science skills, and gradually put myself into the field.
Emily: I wanted a Data Analyst or Data Scientist position. I didn't have a specific company in mind, but I had looked at previous job descriptions at a few companies I was familiar with. I found more job descriptions that required Python than R, so that pushed me towards Metis.
Once you started at Metis, what were your classmates like? Was everyone on the same level as you?
Deepak: For my class, I would say yes. Some people were stronger in programming, while other people were stronger in statistics, but overall I think we were on the same page when we started the program.
Emily: It was quite a diverse group. I had classmates who were my age, a couple of years out of college. We also had people with PhDs in statistics, someone with 20 years of marketing experience, and people with undergrad degrees in computer science. That diversity is one thing that’s really appealing to me about Metis– all these people came to Metis because they were still missing some components.
Emily, could you tell us about your favorite project that you built while you were at Metis?
Emily: My final project was my favorite project. It was sort of “meta,” because I actually did a project about data science freelancers. There's a website called Upwork.com that has lots of freelancing jobs (including data science). I used the Upwork API to gather all the data scientists’ profiles and all of the current jobs posted on the website. My final project was to make a tool that actually matches these data science freelancers to jobs, tailored to their skill set.
That project definitely evolved over time. When I started gathering this data using the API,, I definitely didn't know what the final project would be. As I iterated through it with my instructors and other students, I was trying to figure out "how can I make this as useful and impactful as possible, rather than just give summaries of this data?"
I used Python for most of my project, but I also ended up using R to make the final web application. I used a library called Shiny, which makes it easy to build an interactive dashboard.
Did you find that your final project was important during jobs interviews?
Emily: Definitely. The other great thing about Metis is you also have to write a blog, which I had been meaning to do. When I went to job interviews, I often ended up talking about my project and the challenges. I would give people the link to my application and other projects.
Deepak what was your favorite project?
Deepak: My favorite project was also my final project, because I had a lot of time to work on it, and the topic was something that I was really interested in: soccer statistics. The biggest challenge was getting the proper data and the proper amount of data. I had to reach out to different data scientists across the globe to get the data I wanted, which was interesting.
After talking with the instructors, we ended up expanding the project and creating a soccer betting application. Based on past soccer statistics, users bet on upcoming games and see over time, whether your betting strategy will make you money in the future. I absolutely loved it. The one-on-ones with instructors really helped me understand everything in more detail.
Also, we did presentation practice, which was really helpful and you could see a drastic amount of improvement. And it was all due to the comments and tips from the instructors at Metis, which are the skills that I'm going to carry forward into my job in the future.
What are you up to after graduating, Deepak?
Deepak: I am currently a Business Intelligence Engineer at Facebook. Most of my work is analyzing data, creating pipelines, getting data from different data sources, and visualizing that for our stakeholders. There's just a little bit of data science involved in my job now, but this was a great opportunity for me when it came. I couldn't refuse. I can always build on my data science skills in the future.
Deepak, do you feel like you get to use your skills learned at Metis?
Deepak: I haven't used my skills yet because I'm very new to my job, but there are other people in my team who went to data science bootcamps as well. They definitely say that they use their data science skills, so I'm pretty sure I'm going to use them in future.
Emily, tell us what you are working on and where you're working now.
Emily: I'm on the Analytics team at Etsy, working with about 20 analysts right now, and we essentially embed with other teams. I'm working with our Search team, which is really exciting because search is a huge part of the Etsy experience. Unlike most companies, Etsy’s data scientists are almost all PhDs in machine learning and computer science.
As Analysts, we design and analyze experiments and serve as the “quantitative voice” for our partner teams. Two things I’ve worked on is experimenting with Etsy’s search ranking system and work on opportunity sizing. I hadn't really had experience with big data before, so the learning curve was steep at first. The primary table I use has six billion rows. I've also started using Scala to write Hadoop jobs to get the data that's not stored in a SQL database. That's really exciting.
Amy, are you looking for applicants you think are going to able to land certain types of jobs when they graduate?
Amy: Metis is a vocational training program. Our job is to help students gain the skills they need in 12 weeks to become a data scientist. Most of those positions are entry level data scientists or a data analyst.
We’re definitely trying to identify students who will be successful in the job search after 12 weeks at Metis. 12 weeks is not a lot of time, but the whole idea is that you can adapt quickly when you get into the workforce. We’re looking for passion, communications skills, technical abilities, curiosity, and grit. All these things easily translate into success after the bootcamp.
For data science beginners, what's your advice to prepare for the Metis application?
Deepak: First, understand what data science is. Don't go into the field just because people say it's classy or you make a lot of money. Do your homework, make sure you're passionate about analyzing data and telling stories with data. If you have that passion, and you have basic computer science and statistics skills, I would definitely tell anyone to join the field. There's going to be a huge demand for data scientists in the future.
Emily: There are two sides to this answer. I could recommend specific Coursera courses or books. For example, the John Hopkins Coursera course is great.
The other important prep is to do projects. Don't just read about data science; find a question you're interested in answering with data. There are a lot of data sets that you can basically download. Start playing around with them, as that will force you to learn these skills, and hopefully you’ll enjoy it. If you don’t enjoy it, then this might not be the best path for you. Try to figure out what questions are interesting for you and practice ways of investigating – that would also be great to show in an admissions process. There was a great Quora answer about this subject.
I would be impressed with someone who says, "I don't have a formal background, but I got really interested in data so I developed these predictive modeling skills because I wanted to be able to predict this specific thing that I'm really interested in."
At the same time, don’t be intimidated by data science job descriptions that say you need a PhD in machine learning and have a ton of packages on GitHub. That's not the case. Metis is not meant to take you from 0 to 60, but if you have the curiosity and the grit, don't be afraid to try. Don't feel that just because you don't have a traditional background, you can't make it in this field.
Deepak: Metis will definitely provide you with the tools and skills required for your first data science job. Like Emily said, don't get intimidated. As long as you have those basic skills that we've mentioned and a very keen interest, Metis will provide you with all the skills required.
Amy: I definitely agree with both Deepak and Emily. Do your research, reach out to people on LinkedIn, attend meetups. If this is something you're interested in, then you need to immerse yourself in the world of data science and and really understand the different paths that are out there.
I love when students reach out to me. I'm always willing to have one-on-one conversations to talk about how to move from step A to step B and help them improve in their careers.
Secondly, just do as much coding as possible. Get familiar with Python, do data science projects. Once you get into Metis, you don't want to spend time troubleshooting code. You want to really focus on learning the data science concepts.