Moving to industry from general relativity

[Cygnus_A]

So I'm wondering if anyone has personal or anecdotal experience of people making the specific transition from general relativity to an industry job. It doesn't have to use specific math skills from GR, but bonus points if there is some relation.

It seems that most of the options are somewhere in software engineering or working for the DoD (both of which sound exciting to me). For starters, I'd like to highlight two relevant pieces of advice/experience:

http://www.spsnational.org/cup/profiles/hidden_archives.htm

Elliot Fischer
General Dynamics, Whippany, NJ
"I received a BS in physics and math and a PhD in applied math with a thesis in general relativity. I have spent all of my career (24 years) developing algorithms and systems for the military. These have included target detection and identification algorithms and adaptive algorithms for active control systems. My knowledge of physics has been instrumental in understanding the sources of noise that I have dealt with, as well as how these sources propagate in various mediums. While I usually do not have a direct hands-on experience with hardware, my physics background allows me to understand how certain pieces of hardware work and what their limitations are. In addition, we tend to do a lot of what is called Systems Engineering and Analysis, and again I have found that a general physics background is very beneficial. More recently, we are getting into fiber optic transmission systems, and a background in physics there allows me to quickly pick up new concepts. In summary, I have found that the physics I learned as an undergrad has provided me with a great background to assimilate new material throughout my career."

https://www.physicsforums.com/threads/from-ph-d-to-industry.373844/

Things in industry tend to be extremely field specific, but there are some general tendencies.

One thing that you quickly figure out is not to take job requirements too seriously. I've never seen a job requirement in industry that tells you exactly what the employer is looking for, because often the person that writes the want ad, is someone that really has no idea what the new hire is supposed to do. What I've found works is to spam my resume to anyone that seems to be collecting them, and sometimes someone will call back.

The other thing is that flexibility is important. They might not want you to code FE, but if you can code FE, then presumably you can code lots of other things that have nothing to do with FE.

Yes. Research in a Ph.D. program is usually considered work experience.

Sources of information are alumni that have go on to other things, headhunters and recruiters, people that you know that are in the company that you are interested in.

 

[jedishrfu]

With respect to job requirements, if you write them too specifically then you likely won't find your candidate. If instead you write it more generically then you find people you can interview.

Also teams, will consider how a candidate fits in so sometimes, a good junior programmer joins because an existing programmer takes a more senior position (i.e. the one that was advertised for) and the team is whole again.

Another reason for the genericity is so that a candidate can't complain to a boss saying that's not in my job description because it is. :-)

I was a Physics major who went on to do programming and found that my tensor analysis and understanding the use of indices helped me transition into using array structures in novel ways. You may also get a chance to do some numerical computing as well where your understanding of Diff Eqns and other math will come in handy.

Have you played with MATLAB or Julia or numerical Python?

 

[Cygnus_A]

I've never tried Julia, but I use python and MATLAB all the time. I've been thinking about machine learning as a possible outlet after I finish my phd (if not a professor somewhere -- I've still got lots of time to decide). I spent some time learning about different neural networks and how to code them this summer

 

[jedishrfu]

Look into genetic algorithms too.