This alumni interview was conducted by Liz Eggleston from Course Report.
We asked Jason Liu, a recent graduate of NYC Data Science Academy, all of these questions and more! Read on for his answers, plus advice to other international students hoping to take a bootcamp in the US and how to transition from academia to the fast-paced world of data science.
I finished my PhD in Germany at Ludwig Maximilian university (University of Munich), but I’m originally from Beijing. I was looking for a data scientist position for a while before I started at NYC Data Science Academy. I graduated as a physics Ph.D. last summer (2014). And as a typical physics graduate student, I had no working experience, except research-related job duties for scholarship.
Yes, I have worked with programming for many years. Physics is a hard-core field that requires a lot of technical skills. I self-taught most programming skills and took several data science courses from the Coursera.
Before the Data Science Academy, I had learned Python and other compiled languages like C and C++. I taught myself Python and learned R for about 1 year so I would have exposure to programming. If you learn one language, it’s easy to switch to other languages; there are a ton of packages that you won’t know how to use, but you’ll be able to pick it up.
My goal was to get a senior data scientist role. I wanted to leverage the network at NYC Data Science Academy to gain more exposure to high-end jobs.
There are couple bootcamps in the market. And NYC Data Science Academy is not the oldest. Actually when I applied, that was the first round. However I was impressed by Vivian’s personal passion and visionary insight about the data science industry.
I applied to both the Insight program and the Data Incubator program. However, both programs are funded by job placement. As a foreigner, there is no guarantee for those schools that I could get a job immediately after the bootcamp. So I was not admitted by them. I didn’t consider Metis because it is more focused on programming.
Both. However it was a short and sweet. Basically I went over my CV and answered some questions raised by interviewers.
Yes, the coding challenge is one step in the application process. There are two questions, which could be done in a straight-forward way. However I enjoyed the questions because they are also challenging enough to improve the easy solution for better performance.
The coding challenge was a really open question that we could complete in any language that we wanted. Some of my classmates joked that they did the challenge in SQL, which is a database language that isn’t designed for that type of challenge, but they could finish it. It was flexible, but we needed to prove that we had coding abilities; you could have finished it in Excel, but to do the problem quickly and efficiently, you should use a programming language.
Unfortunately I didn’t get a scholarship. My wife supported me through the bootcamp.
I was in the first bootcamp, which had 14 students. There were 5 Ph.D., a couple fresh graduates, and the rest were working professionals. Just few of them had the same technical background/education as me.
Being a PhD student just means that we had more experience and suffered more in our academic lives. It doesn’t mean that we were smarter, just that we pursued academia. Having a quantitative background was very important. There was another student in my cohort who had a Masters degree and seemed to be one of the more qualified students.
If the person is not from a quantitative background but has learned and practiced a lot on their own, they can be just as qualified.
My instructors are Vivian, Brian, and Jingjing. The class was fast-paced and adaptive. The information was overwhelming at the beginning. However once I got used to the pace, I liked the amount of information and I think the pace helped me to shift from the academia or “research” style (slow) into industry working style.
I was exposed to Programming languages, R and Python. I learned a lot of terminologies for machine learning and skills to use them in applied environment. It was definitely hard to learn them all, so I spent a couple more months continue learning them. However if I was looking for a junior position, I would focus on just couple related skills for my targeted job.
Overall, I like the scope. If there are more time, I would like to learn the current frontier of machine learning.
In my round, there was not exams.
There was a fast feedback loop. Vivian was trying to catch up with comments all the time. I expect to add more Hadoop content with industrial applications.
I worked on an analysis of NYC crime rate. I extracted online information from a NYC government website. I digged into JavaScript/HTML to understand where the raw data is stored. Then I used R to create visualization, hosted a website, and applied exploratory analysis and time series analysis of the crime rate. The project is hosted at https://jiayiliu.shinyapps.io/NYC-Crime-Analysis
I worked by myself and spent 4 weeks on it at the same time took the bootcamp. The most challenging part was to build the project from scratch. At each step of development, I needed to decide which functionality I wanted.
Yes, the academy accompanied with interviewJet to polish my CV and prepare for interviews.
InterviewJet came in to talk about their platform, went over my CV, and helped me turn it into a short, concise paragraph. Through their platform, a lot of employers have access to my profile.
A recruiter reached out to me when I was in the bootcamp for a job at German company Bosch. It took me a while to get the job, and they are still processing my documentation given that I am international.
I’ve already got a job offer at Bosch as a Research Scientists specializing in Big Data. It’s a perfect match, because I’m working with Big Data, plus machine learning using Hadoop and Scala.
We learned Hadoop concept at NYC Data Science Academy. While I am waiting for some government document to start my new career, I’m learning Scala on my own. During my onsite interview at Bosch, I had no problem. Once I learned the concepts at the Academy, I could figure anything out.
There were two rounds of interviews. The first was an HR interview- we talked about my background and communication skills. The second round interview was purely programming. We whiteboarded, they asked questions and I had to write scripts. Because this was a research-oriented position, I also presented my peer-reviewed thesis from my PhD. I had to show my presentation skills.
They asked questions about signal processing, which I had worked with in my Physics degree. When I couldn’t answer questions, they gave me hints and we worked through problems together.
Yes, I am fully prepared for my new company. There are plenty of researchers around and I am looking forward to work with them. My alumni helped me a lot. I learned from them about how to present myself and engage people about my topics.
Typically, you’ll have 1-2 years of working permission after you graduate, so leverage that graduation time in order to take a bootcamp. You don’t want gaps in experience!
Yes, I would recommend spending the money in order to boost your career. For the knowledge part, I would say that you could find all the information by yourselves in this Internet era. However there are still some tips and skills that you might not learn until you learn from instructors and classmates. There’s so much value in the connections and opportunities that couldn’t have been learned by myself.
The first step in becoming a data scientist is to complete your Data Science Bootcamp Application. Just click the button to apply. It's free and will only take you about 5 minutes.