Alumni Spotlight: Andrew Dodd, From Mechanical to Machine Learning Engineer
Andrew Dodd began learning about machine learning on the job as a mechanical engineer. That sparked his interest in deepening his understanding of the field of data science. He pursued that goal by enrolling in the NYC Data Science Academy bootcamp. Now he is applying the skills he learned at the bootcamp as a machine learning engineer at Novetta.
Could you fill us in on your educational and career background before you decided to pursue data science?
I have a bachelor’s degree in mechanical engineering from the University of Massachusetts Amherst and a master’s degree in robotics and space flight from Cornell University. I worked for about a year and a half as a mechanical engineer at Analog Devices in San Diego. I learned about machine learning on the job.
Why did you decide to enroll in a bootcamp?
I wanted to learn more about machine learning but didn’t want to invest the time it would take to complete another graduate program. I figured that the bootcamp would not only be faster but also cost less than enrolling in another master’s degree program.
Why did you select NYC Data Science Academy bootcamp?
I researched which one was best. I read page after page of reviews, starting from best ones downward on sites like SwitchUp. Other bootcamps didn’t compare in the rankings. This one was marked number one on the list because it offers the very strong program in just three months and offers very good job support. I read through the program’s list of topics and found that it had what I was looking for in terms of skills. It not only teaches machine learning but covers both R and Python, which are important for data science. I already know some machine learning but wanted to expand my other data science skills.
How did you find the curriculum?
The curriculum breaks down into statistics with an introduction to R, as well as Python. Then from about week five through week 12, which is all the way through, it goes into machine learning, including R and Python for machine learning. Toward the end, there are two options. You can go into a deep learning focus or a big data focus with different options for lectures in morning or afternoon. You could do either or both. I went to the ones on big data and then watched the live streaming recordings for deep learning for computer vision and other topics.
Aside from the specific languages, which new skills did you learn?
I came in with a background in math and physics, but I didn’t have a lot of business knowledge. A lot of data science is about using data analytics and predictions to drive business decisions. I came in not really understanding how to take the code and translate it into a business application. Those are some of the skills that I gained. Also, I wasn’t so strong in statistics, and this helped me improve in those areas.
How did you find the other students?
My classmates were really great overall. Everyone’s really nice. People get along, for the most part. I enjoyed hanging out with them during happy hour. I really like the fact that you can speak to someone about statistics who has a Ph.D. in stats or in physics. That’s one of the benefits of a cohort made of students with diverse backgrounds. In contrast to a university program where everyone comes in with the same kind of educational background, at the bootcamp, everyone comes in from different areas, which means they have different perspectives on things. There are statistics people, financial people, industry people who have 10 years of experience. There was even an art major in our cohort.
How did you find the projects you worked on during the bootcamp?
We all had core projects to work on. The first project was an interactive platform with R shiny. It involved R shiny and figuring out the best way to visualize the data sets and figure out some business conclusions. The second project involved scraping data from websites. The challenge is that the data is not all on a single page, so you have to figure out a way to pull data from multiple pages. That’s done in Python. The project was interesting because it was an exercise in data science direct to analysis that shows you should buy this rather than that. But the way get data used for that analysis in the first place is often from scraping.
The third project was a machine learning project. I worked with teammates to build at machine learning pipeline to predict housing prices in Iowa using a Kaggle dataset of about two thousand homes with a variety of different features. The project employed lasso regression, XGBoost, random forest, and ensembling techniques.
The last project is a company one. Companies pitch possible projects we can work on. I chose to work with two other teammates. We created a travel search website that allowed customizable search queries. That means that people can click through to get top results for flights and places to stay based on where they want to go and what they want to do. We got the data by scraping 130,000 from Expedia using multiprocessing proxies.
How did you go about your job search, and which NYC Data Science Academy resources did you find helpful?
About a week after bootcamp I went to the bootcamp’s job fair. That gave me a lot of starting leads about where to apply to. I had a few contacts out of that meeting that I then followed up on by email and phone. That gave me three or four leads on jobs. I also looked on job site like LinkedIn, Indeed, etc. for jobs that featured machine learning. I got a lot of interviews in the weeks that followed. During that time I was working as a TA for the bootcamp while applying to jobs and going to interviews. It was really exciting. I’d say it went well.
Can you tell us about your current job?
I’m now working as a machine learning engineer at Novetta. It’s a company that works closely with the government in areas like facial recognition. It also works on a lot of biometric areas like fingerprints and faces. Where I work, in the Tribeca area of New York, we do a lot with facial recognition, so I’m doing a lot with deep learning and image analysis. A lot of these topics I had some background in, and I’m really interested in that area, so it a really good match.
What advice would you have for students considering enrolling in the data science program?
I would start off with talking with someone who knows the program really well and ask what I should get up to speed on before coming because it goes really fast through a lot of topics. It’s good to review areas that you don’t feel super confident in. Also, you can do your own little projects before you come here. What I chose to do is some Kaggle competitions before I came here. I worked on some of the beginner challenges first and then went on to some of the harder ones. That was a really good way to prepare for the program. I would recommend doing the bootcamp’s pre-work. I did it.
It’s a really positive environment. It’s competitive. I liked being surrounded by people all writing code at the same time. I could have learned things by myself, but you’re learning a lot faster when surrounded by others who are doing the same things as you. Everyone wants to do the next cool thing.
This post was first released on SwitchUp.