Wednesday, February 4, 2015

So You Want to Be a Data Scientist?

People reach out to me a lot asking how I got into data science and wanting advice about breaking into the field.  Fortunately this is something I have written and talked about quite a bit!

Below is a compilation of the various things I have "out there" about my transition from academia to data scientist and what it's like to be a data scientist:

What is a Data Scientist? - General overview the answers to some frequently asked questions.

Transitioning Advice / Information
Astronomer to Data Scientist - Advice on how to make the transition from academia to data scientist.
Astronomer to Data Scientist (Talk) - More advice, and some details about what I do for my job.
Astronomy vs. Data Science - Compares and contrasts academia and the tech industry.
Nailing the Data Science Interview - Advice for preparing for your interview and what to expect.
Astronomy to Data Science, Three Years Later - An update on my transition.
Interviews
Interview with Lady Paragons - What my job is like, what I do, some other stuff too.
Podcast with Lady Paragons - More about what it is like to be a data scientist.
Interview with AAS on Women in Astronomy - More about my transition and what my job it like.


Recruiting Advice
Interview with HP - Advice on Where to Find Data Scientists:


Do you have more questions? Ask them below!

2 comments:

  1. Great blog posts! I have found them most enlightening. They do leave me with some important questions, however.

    I have recently received my Ph.D. in astrophysics, and am now looking to transition to data science. In your "Astronomer to Data Scientist" post, you listed specific skills that one will need as a data scientist (familiarity with C++, Python, Hadoop, etc.), and I find that indeed, these are typically listed as requirements in data scientist job postings. I have very few of these, however, having been trained as an astronomer - for goodness' sake, many of us (me!) still program in Fortran!

    I have heard that companies are willing to hire people such as myself who are capable of performing quantitative thinking but who lack these specific skills, and then train us in these skills. If that's true, however, I'm not seeing it reflected in the job postings I've combed through, as they all seem to require that I already possess these skills.

    Is it the case that even though these postings state that these skills are "required", they often aren't actually required, and that the company will hire and train me in these skills? In which case I should apply for these jobs? Or am I performing my search incorrectly? How do I find the jobs where companies are willing to train me?

    You've stated in your posts that the biggest hurdle in this transition is how to convince a company that you can do this job. Can I convince a company without having these skills already, or do I need to develop some of these skills first in my spare time before appying?

    ReplyDelete
  2. Haw,
    Larger companies tend to have the resources to train people on their specific technology stacks whereas smaller companies tend to need people to already have certain technical skills (like SQL, R, Python). If you want to know which companies are open to hiring people straight from academia, I suggest looking at the places where the Insight Data Fellows are employed as these people came directly from academia.

    It is hard to tell from the job description what is actually a "required" skill versus "nice to have." I don't think I've ever had all the "required" skills for any of my industry jobs. Ultimately, there is no harm in applying for jobs, even if you don't have all the required skills, just be honest and let the hiring manager make the call.

    ReplyDelete