After 10 years in the pharmaceutical industry, Tim Chan was ready for a change and saw his future in data science. With a Ph.D. and an M.B.A., the last thing he needed was another degree, so he enrolled in Galvanize’s Data Science Bootcamp in Seattle. After graduating in April 2016, Tim networked like crazy and is now a data scientist for a product analytics team. Tim tells us how he prepared for the Galvanize interview, why the cohort experience was integral to his success, and even shares his final project with us (how much do you know about cheating in the Boston Marathon??).
What first motivated you to change career paths?
I always wanted to design drugs in the pharmaceutical industry, so I got a Ph.D. in Chemistry (and later an MBA). Pharma research is somewhat on the decline, and I found that during my 10 years in that industry, I was gravitating more towards analytics and statistics. Instead of just making the molecules, I was actually using data, correlations, and trends to guide which molecules we should make next.
Once you discovered Data Science as a career, did you think about getting another graduate degree in addition to your M.B.A. and Ph.D.?
Nope, I didn’t consider it. The last thing I need is another degree. When I was evaluating opportunities, the bootcamp made a lot more sense to me. Especially when I heard that Galvanize, in particular, wasn't designed to take students from “0 to 60.” They were looking for students with a solid foundation in math and some programming skills, who they could upgrade into the world of data science. And that seemed a lot more appealing to me.
Did you research other data science bootcamps in Seattle before you chose Galvanize?
I looked far and wide all across the US for Data Science Bootcamps (and I used Course Report)! I didn’t have a straightforward quantitative Ph.D. background, so I didn’t apply for fellowships like Insight. Also, I was looking for a longer program. Galvanize was the right fit for me in terms of curriculum and admissions.
I hear that the Galvanize campus in Seattle is pretty cool.
Yeah. I was living in Boston at the time, and had wanted to move to a West Coast city for a long time. Seattle was high up on my list. And when I found out that Galvanize was actually starting a Data Science program in Seattle, I said "That's it. I'm moving."
I come from the pharmaceutical industry, so this was my first introduction to a tech incubator; it’s got a very cool vibe to it. The Galvanize campus is in Pioneer Square, which is pretty central and fun.
Did you already know some Python or a statistical programming language from your time in pharmaceuticals?
No, not at all. I was pretty into coding starting at ~13 years old, but I hadn’t been coding for over 15 years. I went to a Galvanize Open House before I applied, and they mentioned that they want to see math skills (meaning Linear Algebra, Statistics, and Probability), and Python skills.
I was familiar with all three of those math fields, but I wanted to be certain that I had the depth of knowledge they were looking for, so I took a few online classes.
Which online classes did you take to prepare for the math part of the interview?
I don’t remember the exact numbers, but they were in the MIT Open Courseware. I took an online introductory statistics and a year 2 linear algebra class.
How did you prepare for the Python part of the application?
I really didn't know how difficult the process was going to be. I tried to find information online, tried to hear from other alumni, and tried to find out what the actual questions and scope would be.
Galvanize recommended playing around on HackerRank and being able to do Level Medium questions, which I found quite challenging.
During the Galvanize application process, they actually weigh your math skills and coding skills, and you can be somewhat proficient in one if you at least have a strong background in the other. So they’re looking for a balance between math and programming.
Tell us about your cohort at Galvanize. Was your class large? Was it diverse? Were there a lot of different career and life backgrounds amongst classmates?
After hearing about other Galvanize cohorts, that was actually a big draw for me. I didn’t want to do an online program. I really wanted the cohort experience. That was something that I actually learned a lot from my MBA – I wanted those connections, and I wanted to be able to learn from my classmates. The Galvanize experience definitely gave me all of that and more.
My class was 18 students – roughly a third had Ph.D.s, a third had master’s degrees and a third bachelor's degrees. People came from a wide variety of backgrounds. There were a few people who were fresh out of college and some people who had been working for 10 years. Some people were very strong in coding, while others had a whole career in science or analytics, and they wanted to strengthen their coding skills.
I definitely learned a lot from my cohort, and that was a selling point for Galvanize.
Tell us about a typical day at Galvanize?
In a typical day, we would have a quiz in the morning just to recap previous learnings. Then we had lectures from roughly 9am to 11am. After that, we did a morning coding sprint from 11am to 12am. After lunch, we had another lecture. Then we did an afternoon pair programming exercise where you'd have to team up with a member of your cohort and complete a more complicated exercise to solidify your learnings for the day.
Since you started Galvanize with very little Python skills, how did you get up to speed quickly?
Galvanize has an optional Week 0, and the main focus was on Python and a bit of linear algebra. Week 0 is essentially for people who wanted to get up to speed on the basics. Week 0 helped me start in Week 1 with above-average Python skills, but even in Week 1 and Week 2, I was still learning a lot of Python.
Galvanize promised me that when I graduated, I would be a professional programmer. I didn’t really believe that could be done in 12 weeks, but I definitely feel that now. The Python skills I learned are over and above what I need in my current job. I can talk to other software engineers about Python and feel pretty comfortable in that environment.
Can you tell us about your favorite project that you built at Galvanize?
The capstone project is emphasized strongly, and Galvanize helps students pick projects that reflect their personality. I lived on the Boston Marathon route for 10 years, and while I’m not a runner, I always enjoyed watching the Marathon and hearing the stories of individual runners and their accomplishments. I knew that I wanted to do my capstone project about Boston Marathon data, and I actually used data on prior running histories to figure out who cheats in the Boston Marathon.
Whoa! What did you find? Are people actually cheating in the Boston Marathon??
It's never easy to verify these things. I did get in touch with a blogger, who helped describe what cheating looks like in the Boston Marathon. Runners have to actually qualify, so while they don’t actually cheat in the Boston Marathon, they cheat to get a qualifying time.
Once I learned that, I started to pull people's prior running histories from other marathon events and factor in the course conditions like the temperature and the weather. If I had data from them five years ago, I could do a correction in term of just normal age curve.
Did you use Python for that project?
That project was a lot of Python, and I had to get pretty good with web scraping. Anyone who's written a scraper knows it's actually a lot of work to write one scraper, but to write four of them is four times as more difficult.
Is the project live now? Can we share it with our readers?
Sure, it’s at https://github.com/trchan/boston-marathon.
What are you doing now? Tell us what you do in your new job as a data scientist!
I work in Product Analytics, and essentially I'm the source of data-based knowledge. I scour through the data to answer questions like how is the product doing, where are the problems, and which demographics are enjoying the product.
Do you feel like you learned at Galvanize what you actually need to know for your job as a data scientist today?
There's nothing like actually doing data science for a real company. There's a lot you have to learn in terms of getting up speed with the company. And especially when you're trying to make recommendations to engineering teams- that requires a certain amount of confidence. However, I would say that Galvanize actually prepared me with all of the skills that I needed for my first job. The skills I learned at Galvanize will keep me relevant for many years, and I feel I could perform well in a wide variety of different data science roles.
Anybody who graduates from a bootcamp goes through some level of Imposter Syndrome- how have you built up your confidence as a Data Scientist?
The mentorship that my company provides is pretty solid. I also have a fair amount of work experience, so I've learned how to trust my team with giving me feedback on how I'm doing in the role. I constantly seek feedback, and find that people are pretty honest. You just have to rely on that mentorship and feedback, and continue to grow in the role and just accept the fact that this is a new job, and nobody is going to be perfect on day one.
How did Galvanize prepare you for that job hunt?
We had a hiring day, which actually turned out to not be the most successful event for my cohort in terms of job placements. Previous cohorts had better luck, but for me, I found that Galvanize did prepare me by helping me make sure that my resume was appropriately written like a data scientist resume and with my LinkedIn profile.
Galvanize has an Outcomes Manager who is there to help you with any questions and frustrations you may be having on the job search, networking skills, and setting the bar for a successful job search.
How did you end up getting the job?
I did a lot of networking. I tried to make sure I went to one or two networking events every week. And I know a lot of data scientists will cringe at that because there’s a stereotype that data scientists have an introverted personality type. It's always tough, but you have to push yourself out of your comfort zone. I had three onsite interviews after graduating from Galvanize, and when I look back, all three were from in-person networking.
What is a data science interview like? Are you whiteboarding or coding in Python?
All three of my interviews were pretty similar, but I never had to whiteboard in Python. I did get one short, small coding question, but I did a lot of whiteboard SQL. For a product analytics role, they presented me with a product and a problem, and I had to talk through how I would approach finding the answers in the data.
For example, Twitch may ask you to find the ideal broadcast length. And then you would have to describe to them how you would look in the data to find that answer.
Are there similarities between that world of drug research pharmaceutical business and doing data science work for a tech company? Has your past career been useful to this new role as a data scientist or have you just left that life behind?
I definitely learned a lot of skills as a scientist that are still applicable to my current role. Especially how to define the problem, being absolutely rigorous with your conclusions, making sure that you have accurate data that supports your recommendations or statements. That sounds simple, but you actually learn in the world of science by being burned many times. Also, I learned about using your instinct and following the trail of data in order to get to your ultimate conclusion.
What was the biggest challenge or roadblock in your journey to becoming a Data Scientist? Any advice that you have for people thinking about doing the same type of career change?
For me, the hardest part was actually quitting my job, moving across country and putting all my eggs in this one basket. That wasn't the most intellectually challenging part of the process, but that was the part that I was the most scared about. As a typical scientist, I made sure I had done as much research as I could about that transition, ensuring that the job prospects and salary were indeed in Seattle, and making sure I was taking the right bootcamp.
My advice is to do all of your homework in advance. Make sure you know exactly what you're getting yourself into, and then just go for it.
Have you stayed involved with Galvanize since graduating?
Yeah, I do keep in touch with my alumni network. Like I mentioned before, I think the cohort experience is a very important part of the process, and it pays dividends even after you graduate. You have a network of data scientists who know you well and who can help you out. These are people who you have a bonding experience with, and you've been in the trenches with. To be honest, what I got out of Galvanize was way beyond my expectations going in, so I'm very happy with that experience.