Physics to tech in the PNW?

June 12, 2017 12:54 PM

I've just finished up a PhD doing computational physics, but am looking to move out of academia, because I want to actually decide where I end up living - namely the Pacific Northwest. So I am trying to figure out what I can do, and what sort of jobs I can target, to make myself maximally employable.

I know the 'tech sector' is big here, but that's a vague and broad category, and would like some help narrowing things down a bit so I can target employers where my skills might be more relevant rather than just be good for indicating I'm somebody smart.

The bulk of my work involved writing analysis scripts using python and NumPy, as well as modifying C++ simulation code and running large-scale simulations on high-performance computing clusters. While the code I used was pretty advanced, analysis work I did myself was all embarrassingly parallel, so I only know basic MPI stuff. I'm familiar and comfortable with the basics of programming, classes, data structures, etc. but don't have much knowledge about the newest frameworks or development processes. I know data science is a big thing right now, but I feel my statistics background is a little weak - most of my analysis used things like Fourier transforms and Singular Value Decomposition. That said, I am really quite good at math and can learn quite quickly, it's just that my current recent experience is using vector calculus and linear algebra.

Doing something like a bootcamp that costs money is out of my price range unless I do get an interim job, and I'm ultimately really leery of the commercial operated ones, though I'll consider it if that's the sort of thing that's going to give me a greater than 50/50 shot of earning $75+ k in a year or so, though I get the feeling that's a bit much to expect.

My goal is to get a job that earns $45k+ with decent work/life balance in the shortest amount of time in the greater Seattle or Portland area. Given my skills, what's the best way to do this?

Also, as far as resumes go, what skills do I need to list or highlight to keep my resume from going straight to the trash?

posted by Zalzidrax to Work & Money (12 answers total)

Consider looking into Pacific Northwest National Laboratory. A national laboratory can be a good transition from academia to industry...or you just might stay forever. The work tends to be project-based, so you can do work in many different areas.
posted by answergrape at 1:05 PM on June 12, 2017

National labs tend to look for things like publications and research areas. They also tend to emphasize working in teams and a strong GPA, and may assess your ability to obtain a security clearance.
posted by answergrape at 1:14 PM on June 12, 2017 [1 favorite]

I'll consider it if that's the sort of thing that's going to give me a greater than 50/50 shot of earning $75+ k in a year or so, though I get the feeling that's a bit much to expect.

With a STEM degree (especially a Ph.D) and a bootcamp under your belt you can easily make six figures at a big tech company (Google, Amazon, etc) of which there are many in the region. While these companies don't usually hire people right out of bootcamps, bootcamp grads with existing non-CS STEM degrees is the major exception.
posted by Itaxpica at 1:39 PM on June 12, 2017

I understand that Seattle in particular is quite an expensive city in which to live. I don't think $45K is going to buy you anywhere close to decent work/life balance in either city. Something to keep in mind.
posted by Autumnheart at 2:04 PM on June 12, 2017 [1 favorite]

In general: You may have been creating programs to calculate the weight of a electron (or neutron star) based on some professor's novel theory, or whatever a computational physicist does. However, the job you are looking for may involve you using off-the-shelf software to calculate how much shaving a little weight off a part changes the resulting product. However, they want you because you have the experience and training to know which of the 27 options on the menu are appropriate and what factors can cause the math model to give bad answers. In other words, it's likely you have to drop down a notch or two from the theoretical level you've been operating at. In my first job, the most important skill was Cartesian Geometry.

In specific: Do you like boats at all, and do you have any familiarity with computational fluid dynamics and/or finite element analysis? These things are used in various niches in yacht design and small craft naval architecture. CFD in particular is used for the highest level of racing yachts, and perhaps sail design. The introduction of boats that ride on foils means a move to much higher speeds and many things that have been standardized for years are becoming out of date. Finite Element Analysis is used in designs for all sorts of things, including boats. (Probably buildings, too.) I doubt the usual design office does these things in-house, but they can tell you who the consulting firms are than they use. (Boat designers are, in general, very happy to talk to just about anyone about boat design.) And, I suspect, that new software, and enhancements to software are in the works. If you would like some names, MeMail me.

And by the same token, I bet there are some game companies who would like help programming accurate physics weirdness.
posted by SemiSalt at 4:12 PM on June 12, 2017

Best answer: I'm on the mailing list for The Data Incubator. Basically, it's a data science boot camp for STEM PhDs who are already awesome and just need help tailoring their skills and finding a position. Also, it's free--I'm guessing that they make good money off of job referrals.
posted by hydropsyche at 4:32 PM on June 12, 2017 [2 favorites]

Python coders with physics phd are worth quite a bit more than 45k a year.

You're looking at 200k+ if you get into finance. Update your linked in and start working your contacts and you'll be beating off recruiters with a stick.
posted by empath at 5:15 PM on June 12, 2017 [3 favorites]

And I would repost this question on Hacker News. You'll probably get a job offer in the thread.
posted by empath at 5:17 PM on June 12, 2017 [1 favorite]

you're worth something in the very high five/low six figures, at least, with the right packaging of your skills. really. not all data science jobs are heavy on machine learning; sounds like you could do circles around many people who say they know time series data and filtering methods. highlight that, or if I'm wrong and your thing is more fancy spatial data analysis, highlight that and make sure you show off anything fancy you made in matplotlib, etc. by virtue of having a STEM PhD you are probably better than 99% of people at the fundamentals of probability and hypothesis generation and testing so don't think of your stats background as inadequate; in the business world it isn't especially if you can continue to teach yourself new methods from literature as the need arises, which is what you learned to do in grad school!

you're missing only three things that could prevent you from clinching most entry level data science jobs
1) experience with less structured data and coming up with analysis plans from poorly defined questions other people throw at you - this is just something that takes time and work experience but your technical interviews will probably involve getting a dataset and asking your own questions of it, so you have to be able to demonstrate minimal competence at this. practice on open source data sets that have nothing to do with physics; go play with air traffic data, Census data, EPA water quality data, whatever.
2) some of the software dev fundamentals like git and SQL maybe the Hadoop framework - this is the only thing a bootcamp could teach you really if you already know enough Python. MPI is not common in data science job descriptions but everything Hadoop is
3) business/translational skills - again this is mostly a time and experience thing, but you have to be able to communicate simply and briefly at a level that is less technical than a scientific talk and be able to persuade as you do by constantly connecting back to the value your analysis has for reducing cost or making better decisions or what have you. if you wrote a grant you are capable of this on some level already and it will be a matter of style to adapt to interviews and business clients. this is also something you could maybe get out of a good bootcamp, but these are general enough skills you shouldn't need to pay $10K to polish them
posted by slow graffiti at 6:17 PM on June 12, 2017 [4 favorites]

1. Have you reached out to your advisor and mentors? They have people asking them about possible employees all the time. A personal reference is orders of magnitude more effective than a random application. If you think you have to hide this from them, don't! They already know there aren't enough academia jobs for everyone who graduates, and they will be happy to have an industry contact.

2. Have you looked at Tableau Software? They are all about data, they are hiring like crazy, and they are remote-work friendly.

3. Make contact with some headhunters. It's a win/win/win - they get a commission, you get a job, the employer gets a skilled employee.

4. With your credentials, you don't need to pay anybody for any bootcamps.
posted by metaseeker at 6:57 PM on June 12, 2017

It may not be a good fit for you, but like empath said many people with physics PhDs go into finance as a quant.
posted by procrastination at 7:17 PM on June 12, 2017

I just saw this AAAS webinar in my email and though you might be interested.
posted by hydropsyche at 7:31 AM on June 14, 2017