Many of my former colleagues from my PhD time ask me how to switch from Astrophysics to Data Science. Since I find myself in this transition, I decided to write down my story.
The fairy tale started when I was hired for a PhD position at Leibniz Institut für Astrophysik in Potsdam. Since my Erasmus experience I've been dreaming about a career in Astrophysics and living in Berlin so the dreams were coming true! However, my PhD didn't go as expected, for many reasons, and it ended up with quitting my PhD after 2 years and left me with a tough decision to take: searching for a new PhD somewhere else in the world or staying in Berlin? And I chose for Berlin :)
I started working in Quality Assurance, it is a very easy job to learn for an Astrophysicist and I found it great as first experience in a company. I have to say, In my 3+ years of experience I've learned to work with so many technologies that no Astrophysicist has ever heard about. Just to mention a few, I'm sure you can code in Python, but have you ever used SQL to query relational databases, deployed with Jenkins, built a Docker environment, tested how REST APIs connect several Microservices?
Industry is a total different world from academic world. I'm not saying that one is better than the other, as it depends on each individual's experience and point of view, but they are different. Personally I find working in a company much more dynamic, it taught me a lot about life and about people, and made me become pro-active and organized, all things which are hard to learn in the academic world. In a company there are very tight deadlines on a daily basis, this means the work has to been delivered at latest by yesterday, and you have to learn quickly and get the job done quickly. But don't be scared, working in a company is way much easier than Astrophysics, and it has some advantages: it is never boring, you learn to organize your work better, you learn to prioritize, you learn how to best communicate with people, as communication is fundamental in the business world, you learn how to best sell yourself (which is fundamental for job interviews), you learn to think out of the box...And best of all, at the end of the business day you go home and don't think about work until next morning! ;)
The other side of the medal is that I miss Astrophysics. Ok, I don't miss the academic environment, but I miss doing research, and the big difference with industry is this: while doing your research in Astrophysics, it is YOUR research, your project, and it's something you also do for yourself (at least that was my feeling). In a company you are working for someone else's project and if you don't end up doing some job you really enjoy doing, it could get really tough to motivate yourself in your daily tasks.
1) Consider starting with a Data Science Internship
I know, maybe you want (most likely you need) a full-time position asap! However, think about doing a 3 to 6 months internship in Data Science. Not only it would help you a lot in the transition, but wherever you will send your CV, companies will look at your experience, most of the times exclusively at the experience you already have for the position you are applying for, regardless of how many years you spent in research and how many papers you wrote. However, if you think you might skip this step and apply for a permanent role, try to emphasise your transferable skills, both in your CV and the cover letter. It's plenty of tutorials / articles / videos about how to apply for an industry job when coming from academia.
2) LinkedIn is a MUST
You can connect to business people and build your network. Companies search for your LinkedIn profile when reviewing your job application, they want to see your experience, your network, your recommendations (if you have any). But you also find jobs on LinkedIn and you can receive many offers from recruiters.
3) Coursera and the MOOCs (Massive Open Online Courses)
The two most important courses for starting a career in Data Science, in my opinion: 1. Data Science Specialization by Johns Hopkins University - 10 courses where you will learn fundamentals of Data Science, statistical analysis and R programming. 2. Machine Learning by Andrew Ng at Stanford University, very well explained, very focused on the theory behind the algorithms, lots of algebra (love it!). 3. If you'd like to focus on Python, as many companies use Python, I'm currently having a lot of fun doing Python courses for Data Science on DataCamp!
4) Meetups and Networking
Go to as many Meetups regarding Big Data / Data Science / Machine Learning as you can and do lots of networking. Go and talk to people that count, be nice, take as many contacts as you can and use LinkedIn to connect with them, write emails to those people, meet them for coffee, etc...
Thanks to one of those Meetups I'm now working not only as QA Automation Engineer but also supporting the Data Science team as Data Analyst!
One step closer to Data Science!