alumni-spotlight-arda-kosar-of-new-york-code-and-design-academy

Originally from Turkey, Arda Kosar studied mechatronics engineering, got an MBA and worked as a Business and Sales Consultant before moving to the U.S. in 2015. He had dabbled in data science, and wanted a career as a data scientist, but found his self-training was not enough to land him the roles he wanted. So Arda enrolled at NYC Data Science Academy for their 12-week program to learn R, Python, and machine learning. He tells us why he chose NYC Data Science Academy, how much he learned from his cohort mates, and all about the Kaggle data science competition he entered. Arda graduated in July 2016, and is now a Senior Data Scientist at Publicis North America (and the bass player for a band)!

Q&A

What is your pre-bootcamp story? Your educational background? Your last career path?

I moved to the U.S. from Turkey in October 2015 because my wife got a job here. Back in Turkey I studied mechanical engineering for my bachelor's degree, and I worked for two years as business and sales consultant. Then I decided to pursue an M.B.A., and graduated Jan 2015.

When I first arrived in the U.S., I applied to jobs for three months. I was applying to data analyst and data scientist jobs, but I only received rejections. So I started looking at the job descriptions, to see what skills they required. I saw that the most common requirements were R, Python, SQL, and machine learning. So I decided to make an investment in myself. I didn’t want to do a masters program because they are too long – I needed something short, but efficient. So I found two bootcamps, NYC Data Science Academy and Metis. I decided to start my research on NYCDSA. I talked to a couple of alumni, and they sounded so excited about the bootcamp, and said it was really good. I didn’t need to talk to the other bootcamp, I just decided to enroll in this one. I graduated July 1, and got a job in October.

Why did you want to change career paths and become a data scientist?

In my MBA, I did a course about advanced Excel, and in that we analyzed data, and did some basic machine learning, but at that time I didn’t know it was machine learning. I found that I really liked playing with data and coming up with insights from numbers. So I started looking at data analysis, and data visualization. When I look at a data set, I get so excited about it, and what insights will it give to me. In the projects I did, I found really exciting insights from my data.

Did you try to learn on your own before you thought about a data science bootcamp? What types of resources did you use?

I took some courses from Coursera, but never face-to-face. I knew a bit of Python before I started bootcamp, and I had heard about R but not SQL or machine learning. And I had some basic statistics knowledge, but not that much.

What factors made you choose NYC Data Science Academy?

Their website was so detailed, I saw a lot of things on their curriculum included in the job descriptions I was looking at, and I read reviews on Course Report about NYCDSA. I found it’s really helpful when people leave their name and title on the reviews, so I contacted some of the alumni to ask them about the program.

How did you pay for the NYC Data Science Academy cost?

When I moved here, I had a car in Turkey, so I sold it. I invested the car money into this bootcamp.

What was the NYC Data Science Academy application and interview process like for you?

There was an online application, which asks about your background, and your work experience, plus two coding questions at the end. I don’t know if they are using the coding challenge to eliminate people, but I think they want to see your skills. Then I called NYC Data Science Academy for an interview. I interviewed with Janet and an instructor. They didn’t ask any technical questions, they just asked about my passion about the field. They want to see how dedicated you are, because it’s a huge investment. It’s good because it increases trust, they don’t just say “ok come to the bootcamp”, they are really picky about who they enroll to the bootcamp, which is good.

What was the coding challenge like? What did you have to do and how hard was it?

It was two basic Python questions about palindromes. It had a medium-level difficulty. I had to search a little bit, read about it in some forums, and then I tried to come up with a solution.

How many people were in your cohort? Was your class diverse in terms of gender, race, age, life, and career backgrounds?

We were 20 people and it was a good mix of men and women.  Some of them were managers as I have a friend who is a manager at PWC. One guy was in his 50s and had a son who was our age. It was really great having so many people from all different backgrounds. We could easily ask each other questions, about projects and homework. We were like one huge group that worked together all the time.

NYCDSA usually only accepts people with a master’s degree or Ph.D. Is that something that was important to you? Did you want to learn alongside people with STEM backgrounds?

Yes. I learned a lot from the curriculum, my instructors, and TAs, but I also learned a lot from my cohort mates. It’s really nice being in such a diverse group of people. Some had Ph.D.’s in physics, Ph.D.’s in math, or computer science. I was a little bit scared at the beginning because my bachelor’s degree didn’t include that much statistics – mechanical engineering is just numbers and formulas. I knew some computer science, but I improved my skills a lot with the help of my cohort mates.

What was the learning experience like at your bootcamp — a typical day and teaching style?

When I arrived, before the actual bootcamp started, they put me into an introductory Python class for four weeks, two days a week. So when the bootcamp started it was really nice, because they started from scratch. You know what programming is, but we began with the basics. For example we started with R, we learned how to create variables, write syntax, and other basic stuff. It then steps up really fast, and you have a lot of homework to practice and improve your skills.

On a typical day, the lectures usually start at 9:30am. In the morning there is a three-hour lecture until 12:30pm. Then we have a lunch break. We mostly eat with our cohort, and talk about nondata science stuff. In the afternoon there are no lectures usually, but if they can’t finish the topic in the morning, they can allocate an hour more in the afternoon. Usually in the afternoon, there are homework reviews, coding reviews, or some introductions to different tools to use for our projects, which are useful.

You can stay on campus as long as you want, and you can come in anytime you want. It’s not restricted – you can even sleep there! I didn’t sleep there, I live in the Bronx, so it took me an hour and 15 minutes to commute. I was, however, the first person in, and the last person out most of the time. At the end of the bootcamp, most of us stayed there til 7pm or 8pm, but it really depends. If you’re more efficient working from home, you can go home, but generally all the material the instructors and TAs taught ended around 3:30pm or 4pm. There were also sometimes guest speakers from the industry, which was really cool. Usually the day ends at 4pm or 5pm.

What were your Instructors like?

We had an instructor teaching R, machine learning, and statistics who has a really powerful background. We also had another instructor who is teaching Unix, Git and GitHub, and creating Shiny dashboards in R. Another instructor was teaching Python, and machine learning in Python. We also had three or four TAs. So there were a lot of people to turn to whom we could ask questions. They were super helpful. You also have a Slack channel so you can Slack them if you’re not on campus.

What is your favorite project that you worked on at NYCDSA?

I’d have to say my capstone project. We worked on a Kaggle competition with a group of three. It was an open Kaggle competition, so every day there were neo-calls and neo improvements, but we got 30th place. It was about predicting demand from historical sales data. It was a nice project and really business-like. It was for a Mexican bakery company, and was a really common business problem, about reducing the amount of leftovers. It was really nice to see that machine learning can be applied to real business problems like this. It was a real bakery  called Groupo Bimbo.

Kaggle is a platform for open data science challenges, so they are real business problems, but with simulated data sets, because they don’t want to share their actual numbers. But the problems are real, the size of data sets are real, only the numbers are simulated. It’s open to everyone around the world. One cool thing about Kaggle is there are some really experienced data scientists competing, so even by reading the forums and looking at their solutions, you can learn so much.

How did the bootcamp prepare you for job hunting (interviews, hiring events, whiteboarding)?

NYCDSA works with a resume review company, and each student gets a one-on-one resume review session. Vivian, Founder and CTO of NYCDSA, and two other people on the job hunting team, are helping students a lot, they are continuously watching your application process, and looking in their networks to see if they know people at the companies you are applying to. NYCDSA also does mock interviews, so they are doing their best to get you a job. To actually get a job depends 30% on their effort, and 70% on your effort. They are not magicians but they are doing their job really well.

What are you doing now? Tell us about your new job!

I work as a Senior Data Scientist at Publicis North America, a marketing agency. For now, I mainly build tools to help other departments. It’s only my second week, so I haven’t used many machine learning techniques I used at the bootcamp, but I’m using all the information I learned about R and Python, so now mostly it’s about programming. I think in the future there will be a lot of new stuff. The tools are mostly visualizations for other departments if they have to prove something. So if you have to prove something to someone, you have to come up with data, or visualizations. Some of it is to simplify their workload.

How did you get the job?

I applied to the job through LinkedIn, but actually when I got the onsite interview, I contacted Vivian to tell her I got an interview at Publicis, and I discovered that just a week ago my director and my manager gave a speech at NYCDSA. So they already knew Vivian and NYC Data Science Academy which made my life easier.

What was that interview process like for Publicis?

I had a phone interview first, to check if I’m a good fit for the job. Then I had an onsite interview which took about half an hour, and I talked about my projects and my MBA. They asked a lot of questions about my projects. I didn’t have to do a technical interview with coding challenges or data sets, but they asked technical questions about my projects.

Has your previous background been useful in your new job?

Yes. In my MBA I mainly focused on marketing, which helps a lot because Publicis is a marketing agency. So I know the terms and the jargon, so if you combine it with a data science bootcamp like this, I think it’s a good match for marketing agencies.

I also took some programming classes in my engineering background, and some basic statistics. But other than that the only thing I’m using is how to think mathematically, how to troubleshoot, how to think like an engineer. So I’m not using the mechanical engineering classes.

What’s been the biggest challenge or roadblock in your journey to becoming a data scientist?

I didn’t get that many math classes in my bachelor’s degree, so the most challenging thing was catching up with that stuff, because machine learning is on a basis of mathematics. So I had to catch up with work and homework, doing the projects. So that’s why I was the first person in, and the last person out most of the time. It was the most challenging thing for me just to keep up with everyone, because you also have to get some sleep and take care of yourself.

You have to really dedicate three months of your life just to this, and nothing else. I was still able to do rehearsals and play shows with my band, so that was my only social stuff. I’ve been in a band called Tacoma Narrows since February. We have an album on Spotify. It’s Folk Americana, and a little bit of funk. It’s a mix of genres. I’m playing bass. You have to have something to clear your mind from your huge workload.

How do you stay involved with NYCDSA? Have you kept in touch with other alumni?

We have a Slack channel for all of our alumni. We are going to alumni events, and we are also organizing reunions among ourselves. We talk to each other all the time. It was only 20 people, so we still keep in touch, meet, catch up, have some drinks.

What advice do you have for people making a career change through a data science bootcamp?

You should only expect 30% from the bootcamp, it’s all about hard work and dedication. You really have to dedicate three months of your life to this and most of the time you won’t be able to do anything else. That’s the key I think. If you are doing a career change like me, you really have to make sure your resume and LinkedIn are polished, because if you are applying online, they are the only things that represent you. These things have to be really solid, and they have to show you can do data science. Vivian is helping a lot in that process! We also post all of our projects online in a blog like a portfolio. You can see my portfolio here.

Find out more and read NYC Data Science Academy reviews on Course Report. Check out the NYC Data Science Academy website.

About The Author

Imogen crispe headshot

Imogen is a writer and content producer who loves writing about technology and education. Her background is in journalism, writing for newspapers and news websites. She grew up in England, Dubai and New Zealand, and now lives in Brooklyn, NY.