Seeking help moving from academia (Econ prof, Math PhD) to industry (Data Sci.)

[davidmayerf]

Hi, My name is David. I am Mathematics PhD ( received my Bachelor's and Master's degree from Oxford and my doctoral degree from UNAM in Mexico) and I am currently an Economics professor. Over the last couple of years, I have become increasingly interested in Data Science and Machine Learning and I would like to make the transition to industry work. I have spent most of my professional career as an academic, and would really appreciate any advice on making this transition!

[qwrusz]

Hi David,

I made a similar move to industry work many years ago. My advice is to clarify what interests you, as data science and machine learning are huge fields. Also "machine learning" is still hard to pin down what that even means these days. If you're willing to start off in a data science role without a heavy machine learning aspect to it, that's very easy; there's tons of companies (startups like Uber and more mature companies like Facebook) that have entire nap rooms of data scientists with PhD's working there and some super interesting data to work with.

If there is a specific area of data science and/or machine learning or other work going on that interests you that you come across, email a PhD at the company doing this work. Many of these people still publish papers at conferences or at the very least have blogs describing what they are working on. Find a company that's doing work that interests you and email the researcher. Don't go through HR. Email the person directly and ask to chat. I found PhD's working in industry to be very receptive to talk to someone qualified and with an interest in the work as they have likely been in your shoes before. Go from there. GL!

[davidmayerf]

Thanks for your reply! It is very down to earth and to the point.

[rchaud]

A friend recently recommended the 7-week tuition-free Data Science intensive program at Insight Data Science (http://insightdatascience.com/).

The program is specifically for PhDs who are in your position, aka moving from academia to industry. I believe the course is available in Boston, New York and remotely (not 100% sure of the latter).

I'm in a similar boat. I studied Economics in undergrad, but I only have a Masters (in an unrelated field) so I didn't meet the critieria.

[jonhurlock]

First step apply. With that background, you have a high chance of getting a job straight off.

The problem you may face, is you will need to know how to program in a 'production' environment. So were as in academia you can be a single person running code, in industry, you will have multiple people utilise and check your code. It will have to be robust and scalable. So you may need to brush up on programming skills, though you could be great from the start.

Something I've seen with Academics, is that whilst they are great at solving detailed problems, they sometimes fail to present the impact of their work in a business context, or what they can offer and how it solves business problems, rather than just "interesting" problems.

My advice would be to apply and see what comes back, and get feedback from the people you apply to.

[davidmayerf]

Thanks! That is more or less what I am doing

[Rainymood]

Interesting! How did you go from Math PhD to econ professor? I doubt you just applied to some econ tenure position. Please elaborate, I am very curious about your history!

The reason I am asking this is because I'm enrolled in an econ PhD myself (spec. in advanced econometrics) and I am extremely interested in machine learning and data science. I am constantly looking for ways to combine the two in interesting ways. It's nice to meet someone like-minded.

[davidmayerf]

Send me your e-mail, I'll get in touch. I hope to keep on doing Econ with machine learning whatever happens. I did just apply to an econ tenure position, but it was not in the US.

[Rainymood]

I am not entirely sure how to private message someone on HN so I will temporarily put my email in my profile.

[davidmayerf]

Thanks, got it, I'll write soon

[pjmorris]

Can you find an industry collaborator to work with on a research project? I'd wager someone in your department is willing to bend your ear about interdisciplinary collaborations. Finding (or proposing) something interesting, then applying data science techniques could yield learning, funding, and publications. Rinse and repeat until you've got the skills you want or the new job you want.

[maxmoo]

can i have your old job? i've been in industry for last couple of years after finishing my phd, and my job feels futile and meaningless (btw yes i'm good at solving "business problems", but i feel pretty depressed/terrified if this is going to be my lasting contribution to the world)

[davidmayerf]

Is your PhD in Economics? Maybe there would be a place for you where I work.

[pravenj]

Could collaborate on a project that I am contemplating as well as testing out bits and pieces. One thing though I probably am located very far from you and the project would be more of a research project. If interested a few rounds of emails might become a path to follow.

[davidmayerf]

Write me at [e-mail]

[itamarst]

Beyond other suggestions there's things like http://insightdatascience.com/, which is specifically aimed at PhDs.

[davidmayerf]

Great suggestion!

[pgroot]

just do what you want to do the best you can do, and learn as you go. mistakes happen and you learn from them. There are plenty of academics who have moved to industry in your field, may be they can chime in. but you should just apply, and go from there.

[davidmayerf]

OK, great, that is what I am trying!