Near the End of A PhD and Have No Job

[twofish-quant]

If these guys can do it, everybody else with a physics degree can.

Not everyone. If you think investment banks are fundamentally evil, then this disqualifies you. If you have no knowledge in finance, that's a handicap (albeit a rather minor one). If you have no *interest* in finance, that's the kiss of death. Same for programming. It's not fatal if you have very, very basic C++. However, if you have no *interest* in C++, this is going to kill you.

There's also the "you have to move to NYC/London/HK/Singapore" condition. The thing about finance cities is that they are "type A" cities. NYC is a city filled with hyper-ambitious people all out to make it big, and everyone is out to make a buck. Some people hate it. I fell in love with the city.

 

[Mépris]

Not everyone. If you think investment banks are fundamentally evil, then this disqualifies you. If you have no knowledge in finance, that's a handicap (albeit a rather minor one). If you have no *interest* in finance, that's the kiss of death. Same for programming. It's not fatal if you have very, very basic C++. However, if you have no *interest* in C++, this is going to kill you.

And who's fault is that? :)

I said everyone with a Physics degree *can*, not want.

There's also the "you have to move to NYC/London/HK/Singapore" condition. The thing about finance cities is that they are "type A" cities. NYC is a city filled with hyper-ambitious people all out to make it big, and everyone is out to make a buck. Some people hate it. I fell in love with the city.

Sounds cool. :):)

How do you do that if you're a foreigner?

 

[jk]

And who's fault is that? :)
How do you do that if you're a foreigner?

The banks don't care as long as you have the legal right to work in the country.

 

[mark55]

Wow, my thread blew up while I was away!



Generally I get some lip service about how I had impressive qualifications and such, but "just don't fit any of their openings right now". Generally these people will point out that I do have qualifications that do fit with them, just not with anything at the moment.



This is what I do. You get the general sales pitch, and then you get asked, "so what do you do?" and no matter how you try to sell "physics", they laugh and say "sorry, I don't know how to help you. Apply online." It never, ever, ever works. Ever.



Yes, but my network is poor, as I have said. Most of the time when I send emails to random people, I get ignored.



Actually, that's *exactly* how it is designed. It's designed in a way that if you went into grad school, you're completely and totally screwed. This is because you are now too qualified for entry jobs, and have not enough specific experience for the higher levels.

As an example, I applied to many, many entry level jobs that merely required bachelor's degrees in various technical fields. I usually got rejected immediately and when I asked why, the few responses I got back were "PhDs aren't entry level". When you look at the higher levels, they typically demand you know something that either requires Yankee White security clearance or 5 years doing something very specific to that company or industry. You have no chance.

The system is absolutely designed to screw PhDs.



Not to be rude, because I know you're trying to help, but I don't know how many times I can say I go to conferences all the time. It doesn't help. You always just hear "go to the website" or the contacts will offer to pass ur resume around if they like u, and you don't hear back. I'm not taking this personally, but it's just not helpful.



This attitude is devastating, especially as it's coming from someone who used to do interviews. It's absolutely maddening that something that is even recognized as "learn able in a month" is considered beyond our grasp.

I run into this with C++. I know Python very well, as I coded in it my entire grad school career. I've been dabbling a little with C++ as per two-fish's recommendation, but not quite enough to really put C++ on a resume. I absolutely have talked with people that say "heh, you haven't done OOP? we'll get back to you." It's just stupid to me that they absolutely recognize that i have the required skills they want, but just because I don' have one LITTLE thing (such as coding paradigms, OOP isn't difficult), they literally laugh.



But this is horse manure. It would take all of a month to catch up to speed, especially if you having coding experience with another language.

I don't think so, unless you have serious natural ability - c++ is a tough language to get good at. You can learn to do basic stuff in a month - but I don't think many people can be any good at it (especially given the poor code quality I see from people who have used it for some time!). All my IT friends say it has a fearsome reputation in their circles and my fluency would stand me in good stead because not many people are good at this language due to its difficulty.

python uses OOP anyway (amongst other programming methodologies), so you don't need to know anything about c++ to understand OOP and have experience of using it. So you could learn it without touching c++?

A friend of mine had trouble breaking out of academia into software - writing his own software that is used by a local social group really helped him get a job because he could show this software off and say its good enough quality that people are using it. So that's an option (if you want to break into programming).

 

[jk]

I don't think so, unless you have serious natural ability - c++ is a tough language to get good at. You can learn to do basic stuff in a month - but I don't think many people can be any good at it (especially given the poor code quality I see from people who have used it for some time!). All my IT friends say it has a fearsome reputation in their circles and my fluency would stand me in good stead because not many people are good at this language due to its difficulty.

python uses OOP anyway (amongst other programming methodologies), so you don't need to know anything about c++ to understand OOP and have experience of using it. So you could learn it without touching c++?

A friend of mine had trouble breaking out of academia into software - writing his own software that is used by a local social group really helped him get a job because he could show this software off and say its good enough quality that people are using it. So that's an option (if you want to break into programming).

I agree with this. There is a sort of arrogance that says that you can learn a programming language in a few weeks. While this is true, learning a programming language (i.e the syntax) is not the same as learning to become a good programmer. From my experience, it usually takes about 2 years of working as a programmer to become decent at it. Even then, I would consider that as Junior to Intermediate for resume filtering purposes until I meet the person and have had a chance to interview him/her. I used to do interviews for a financial company and I found this to be true for most people.

Programming in the environments I have worked at involves not only being able to write code that compiles and runs but a myriad of other skills that take time and experience to learn: design patterns, debugging skills, performance and scalability considerations, knowledge of the common data structures and algorithms along with their applicability to different situations, knowledge of different vendors'/open source products, understanding of threading and parallel processing concepts and idioms, SDLC, documentation, some project management, presentation skills...there is a whole host of knowledge that goes into becoming a good programmer. 30 days with "C++ for Dummies" will not give you that experience and knowledge.

 

[twofish-quant]

I don't think so, unless you have serious natural ability - c++ is a tough language to get good at.

For a lot of jobs, you don't have to be good. You just merely have to be not incompetent. C++ is a very rich and deep language. I've been programming in it for 20 years, and I'm still learning lots of new stuff, and I don't consider myself an expert in it.

But you have to work with what you have. If you don't know Armenian, then you aren't going to be writing novels in a month. In a month I think you *can* get yourself to the point were you can ask for directions and survive. Same with C++, you aren't going to be an expert in a month, but most Ph.D.'s can get to the point were they can be useful with it.

All my IT friends say it has a fearsome reputation in their circles and my fluency would stand me in good stead because not many people are good at this language due to its difficulty.

The difficulty in C++ is that it's multi-paradigm. You can program C++ that looks like C. You can program C++ that looks like lisp. You can program C++ that looks like java. You can program C++ that looks like Fortran.

The reason that C++ is so widely used is that it's the Swiss army knife of languages. If you have a situation in which you have to mix Java and Fortran programming styles, there aren't any other options.

So don't try being an expert in C++. If you are a Ph.D., you figure out enough of C++ so that you can program code that looks like fortran.

python uses OOP anyway (amongst other programming methodologies), so you don't need to know anything about c++ to understand OOP and have experience of using it. So you could learn it without touching c++?

You could but what's the point?

Except that this kills you if you are looking for jobs that have systems that are written in C++.

A friend of mine had trouble breaking out of academia into software - writing his own software that is used by a local social group really helped him get a job because he could show this software off and say its good enough quality that people are using it. So that's an option (if you want to break into programming).

This gets to the attitude thing. If you can only program basic C++, this doesn't kill you. The important thing is not how much you know, but how much you are willing and able to learn, so programming some real world software helps a lot. Also you'll learn about the politics of software.

 

[twofish-quant]

You need experience to get a job, and a job to get experience. Chicken meet egg.

While this is true, learning a programming language (i.e the syntax) is not the same as learning to become a good programmer. From my experience, it usually takes about 2 years of working as a programmer to become decent at it.

Once you are a skilled programmer at one language, you can move to another programming language very quickly. If you can write decent fortran 77 code, then switching so that you can write decent C++ code isn't that hard.

There is a whole host of knowledge that goes into becoming a good programmer. 30 days with "C++ for Dummies" will not give you that experience and knowledge.

It won't but that's not the point.

Also one reason that theoretical physicists are hired is that they often have experience in a different area. Numerical programming is a separate subfield and the skills are different. A month of self-teaching won't make you an expert C++ programmer, but it will get you enough so that you can use your numerical programming experience in a C++ environment.

 

[twofish-quant]

And who's fault is that? :)

I said everyone with a Physics degree *can*, not want.

Fault doesn't matter, and it can be harder to change one's fundamental beliefs and personality than to change one's skills. For me it took me a while (several years in fact) to fall in love with finance and the NYC lifestyle.

I know people that have physics/math degrees that just will not fit in an investment banking environment because of personality or attitude reasons.

How do you do that if you're a foreigner?

Get a Ph.D. and while you are in school start looking for jobs. It's hard, maybe impossible, to look for work from another country, and one thing about universities is that it gets foreigners into the country so that they can see what jobs are available.

 

[mark55]

You could but what's the point?

Except that this kills you if you are looking for jobs that have systems that are written in C++.

I thought the poster said he was not getting jobs because he did not know OOP, but did know python. What I meant was you don't then need to learn c++, if you know python already - learn OOP there and you know how to do it. Now if the job wants c++ AND OOP that's a different matter.

 

[mark55]

You need experience to get a job, and a job to get experience. Chicken meet egg.
Once you are a skilled programmer at one language, you can move to another programming language very quickly. If you can write decent fortran 77 code, then switching so that you can write decent C++ code isn't that hard.

I'd disagree I think because generally c++ is using programming techniques not used in FORTRAN - most code (at least that I come across) in FORTRAN is procedural and most code in c++ has some aspect of OOP in it. So sure you can write procedural code in c++, but I have rarely come across code in c++ that is not using OOP ideas. So there is something new to be learned to make the switch.

When I first used python I ended up just writing c++ code in python syntax, it took a while (and begging my boss to pay for me to attend a course) before I realized there is a lot of stuff in python I can learn about which does not exist in c++. And learning about that new stuff is pretty useful to write better code in python. So I could write code in python that would work quickly enough, but I would not say I was much more than a python beginner being able to do that, even though I had already many years experience in c++ and consider that I know my way around it very well.

I suppose in the end its what the employer wants - maybe some only care you can cobble some code together that will just about work (and that can even be ok for certain applications - e.g. analysing data for your job and only you or a few others are using it), but others might care you really know your way around and write very nice code (e.g. has to be sold to lots of people so it cannot have even obscure bugs that cause occasional problems because you want to get a good reputation with your customers for the products you sell, needs to be easy to maintain for many years in the future by people that come along after you leave for new jobs etc).

 

[twofish-quant]

I'd disagree I think because generally c++ is using programming techniques not used in FORTRAN - most code (at least that I come across) in FORTRAN is procedural and most code in c++ has some aspect of OOP in it. So sure you can write procedural code in c++, but I have rarely come across code in c++ that is not using OOP ideas.

Remember that the focus of this discussion is "What do I as a Ph.D. have to do in order to get a job?" and the focus of what I'm trying to do is "what you have to do so that I'm not forced to toss your resume immediately."

The way that a lot of physicists "adapt" to C++ is that you have this huge nasty system that is written in C++ and there is a piece in which the Fortran-trained physicist/programmer is told to "do your magic here." In order to get to that level of programming skill will take you one to two months. It will take you a year to be a proficient programmer in C++, and maybe a decade to be a C++ guru.

However, you don't have a decade, so what I'm trying to do is to tell you what is the bare minimum you have to learn so that I can do something with your resume. If you are a skilled fortran programmer, and you know just enough C++ so that I can tell you "please fill in the blank here", then that's enough for you to get hired. Once you get hired, then you can learn the rest on the job.

Note here that for this to work, it has to be a job in which your fortran skills are considered useful. This is *NOT* true with most programming jobs, but it happens to be true in jobs that require lots of PDE crunching. If you have no idea what template metaprogramming is but you do know what the Courant and Von Neumann stability conditions are, this is useful for what I'm doing. If you have no idea what a virtual function is, I can't do anything with your resume.

When I first used python I ended up just writing c++ code in python syntax, it took a while (and begging my boss to pay for me to attend a course) before I realized there is a lot of stuff in python I can learn about which does not exist in c++.

And there is a lot of cool numerical stuff that you can do in C++ that you can't do in fortran. Once you get hardcore into numerical C++ programming you will get into the mysteries of template metaprogramming, and look at Boost::Math.

But it helps to learn the cool new stuff if someone is paying you to do it.

you want to get a good reputation with your customers for the products you sell, needs to be easy to maintain for many years in the future by people that come along after you leave for new jobs etc).

One reason that fortran experience is useful is that the type of bugs that you find in numerical code are things that physics Ph.D.'s have spend years tracking down. Also it turns out that maintainability isn't the most important thing for financial code. Not losing a billion dollars and getting your company in Bloomberg in a bad way is.

 

[whyee]

Personally, I find that optimism doesn't work that well.

Cynicism and a taste for the absurd works better for me. Also, I found that looking for work left me profoundly angry, and a lot of the "how not to go too crazy" was to deal with the anger. It turns out that for me, anger was useful. The thing that I had to worry about most was getting so depressed that I couldn't get out of bed, but the nice thing about getting angry was that getting angry gets you out of bed.

Reminds me of a quote I heard (had to do a search to find the exact quote and source): "That of course is the advantage of being a pessimist; a pessimist gets nothing but pleasant surprises, an optimist nothing but unpleasant," which is from Fer-de-Lance by Nero Wolfe.

 

[stevencruiser]

Where there is a will,there is a way.Be optimistic and fight with the difficulties.

 

[mal4mac]

Just keep on applying! And don't dismiss Academia... I'd expect you to have a chance to get "Research Software Engineer" kind of jobs in Academia with just python. It would help if you could put a compiled language on that CV. Have you never taken a course in C, Pascal or the like? Academia will rate your PhD far more highly than industry if you are competing for jobs that only require a BSc. If you are applying in the UK, check out:

http://www.jobs.ac.uk

... just type python and see what happens. I'm not sure if there is a similar site for other countries, maybe others can chip in with suggestions.

Also, why not apply for a teaching training course, at least as a backup plan! Again, I'm not too sure about the situation elsewhere, but in the UK you could do a PGCE and get a grant for doing that. If you don't fancy coping with unmotivated schoolkids forever there's always the chance of eventually "teaching the teachers" by moving to a teacher training college lecturing post... and a PhD + PGCE will certainly help for that move...

 

[mal4mac]

If you are a skilled fortran programmer, and you know just enough C++ so that I can tell you "please fill in the blank here", then that's enough for you to get hired. Once you get hired, then you can learn the rest on the job.

Note here that for this to work, it has to be a job in which your fortran skills are considered useful. This is *NOT* true with most programming jobs, but it happens to be true in jobs that require lots of PDE crunching.

Nah... you can get any kind of job with Fortran experience, at least I did. I did heavyweight PDE crunching in Algol and Fortran and next job went on to develop human computer interfaces/expert systems in Basic (of all things :) The interviewing prof. was a physicist who had gone the same route ... all you need to do is to be very confident at interview and convince the prof that you are good at picking up new things, like him...

 

[Locrian]

Nah... you can get any kind of job with Fortran experience, at least I did.

Really? You got any kind of job?

Or did you get two jobs?

 

[twofish-quant]

Where there is a will,there is a way.

Sometimes there isn't. One thing that you have to do from time to time is to give up and try something else.

Be optimistic and fight with the difficulties.

Optimism can be a bad thing. Look up Stockdale paradox.

 

[mal4mac]

Really? You got any kind of job?

Or did you get two jobs?

More like seven jobs, all using different languages in different application areas, some internal transfers - but still makes the point, *any* programming skills are readily transferable to other languages and areas.

Maybe not *any* job, but t-q was implying that Fortran skills were a narrower qualification than they are in actuality. Maybe because he has had only one job? He said: "it has to be a job in which your fortran skills are considered useful. This is *NOT* true with most programming jobs, but it happens to be true in jobs that require lots of PDE crunching."

I think *IT IS TRUE* with most programming jobs. For instance, you can transfer to Basic or C in fairly short order. In one transfer, I was employed to write Object Pascal programs with only Fortran/Basic/C (no OOP) experience.

If you have the skills to write heavyweight PDE programs, you have skills that can be transferred to writing many kinds of program. As Zapperz stressed, you might have to, also, do a lot of learning on the job. Learning OOP properly from scratch was a big job! But my Fortran skills got me in the door...

 

[ParticleGrl]

Maybe not *any* job, but t-q was implying that Fortran skills were a narrower qualification than they are in actuality.

In my experience in the US, you won't even get an interview with most programming jobs unless you specifically list the language they are looking for on your resume. In that sense- Fortran is quite a bit narrower than most languages. HR isn't going to say "well, if he can do numerical programming in fortran..." they are going to say "i was told to pass up only those resumes that say java."

 

[mal4mac]

In my experience in the US, you won't even get an interview with most programming jobs unless you specifically list the language they are looking for on your resume. In that sense- Fortran is quite a bit narrower than most languages. HR isn't going to say "well, if he can do numerical programming in fortran..." they are going to say "i was told to pass up only those resumes that say java."

It's a similar situation for many non-academic appointments in the UK. It comes from HR departments not being flexible enough. But good professors know that bright students can easily pick up, say, Basic if they know Fortran, and they can easily develop interfaces if they can program PDEs.

I'm not sure how universities work in the states, maybe the HR filter is more "up front" than in the UK. But, surely, professors looking to (say) program PDEs in C++ would stress to HR departments that numerical Fortran programmers are acceptable! (Or is there such a glut of unemployed C++ programmers this year that Fortran programmers are heading for skid row? Is it really that bad out there?)

"Working in a bar" doesn't look great on the CV - good for keeping people skills polished, but some evidence of continuing brain work would be useful! Why not do some open source development? Or volunteer - charities may be looking for people to develop databases or websites... Keep the brain working and add some languages to the CV...

 

[Astro_Dude]

Update -

I was fortunate enough that I was able to find some temporary work between my graduation and the end of the year to keep my *** afloat. I applied for hundreds upon hundreds upon hundreds of jobs, and never got anything past phone interviews.

In the end, I got a job (non-permanent) doing much the same work I did as a grad. It pays ok, and I'm at least enjoying the environment. I have to count whatever blessings I have though I guess until I can manage something better.

The fact is, industry has no interest in me, and I'm going to try to learn a compiling language while I'm here. It's the only real way to get more interesting to them. Also, maybe there will be a change in Washington to actually make hiring in industry actually happen again.

I want to thank everyone for their support here through this dark time. I am at least going to survive for a *little* while...

-AD

I'm not sure how universities work in the states, maybe the HR filter is more "up front" than in the UK. But, surely, professors looking to (say) program PDEs in C++ would stress to HR departments that numerical Fortran programmers are acceptable! (Or is there such a glut of unemployed C++ programmers this year that Fortran programmers are heading for skid row? Is it really that bad out there?)

Here's how it works.

HR: Well, you're clearly smart enough to handle this position, tell us why you want to change fields. Ok, *other questions*. Sounds good, I'll pass this along to the next person, you should hear back in about two weeks.

OPTION A:
*it was a lie, not passed along*

OPTION B:
*passed along*
"Hiring Manager" : Oh, well, screw this person, they can't hit the ground running immediately. NEXT

------

There's a huge glut of programmers. You have to be extraordinary if that is your only skill. The thing we *do* have an advantage on is we can do the scientific programming that a lot of industry needs. That is, a lot of CSCI majors apparently don't have the math skills physics majors do, and certainly don't have the physics understanding we do. So it will help if you want to go into coding a missile or something.

 

[jk]

In my experience in the US, you won't even get an interview with most programming jobs unless you specifically list the language they are looking for on your resume. In that sense- Fortran is quite a bit narrower than most languages. HR isn't going to say "well, if he can do numerical programming in fortran..." they are going to say "i was told to pass up only those resumes that say java."

While this is true in a lot of cases, it also depends on the company and their needs. I once got a job because the hiring manager was convinced that I could learn the skills necessary to be a developer even though my work history did not include development.
Another example: I just got a gig working with the Ruby programming language - I know zero Ruby at the moment. The hiring manager was convinced by my history, references and whiteboard interview that I can be productive in Ruby fairly quickly

 

[Astronuc]

I'd disagree I think because generally c++ is using programming techniques not used in FORTRAN - most code (at least that I come across) in FORTRAN is procedural and most code in c++ has some aspect of OOP in it. So sure you can write procedural code in c++, but I have rarely come across code in c++ that is not using OOP ideas. So there is something new to be learned to make the switch.

Fortran was adapted to OOP in 1990.
http://www.clear.rice.edu/mech517/F90_docs/EC_oop_f90.pdf
http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/21636/1/97-0004.pdf

There is some evidence to suggest that OOP Fortran is more efficient than C++.

 

[ParticleGrl]

Fortran was adapted to OOP in 1990.
http://www.clear.rice.edu/mech517/F9...EC_oop_f90.pdf
http://trs-new.jpl.nasa.gov/dspace/b.../1/97-0004.pdf

One caveat- most of the people I know using fortran code (which is academic HEP) are using it for legacy reasons, so its big chunks of fortran 77, not fortran 90. So its not just that they are coding fortran, they are coding on obsolete fortran.

 

[bpatrick]

I know this might not be of any help to you, since I pretty much browsed the thread, but I have a few friends who have had horrible luck coming out of their PhDs trying to get jobs in applied math, astrophysics, and theoretical physics.

Becca ended up taking a year off and spending 80+ hours a week learning stats/finance stuff and passing 3 actuarial exams in her first year post PhD (PDE theory was her dissertation subject). She's now doing well as an actuary somewhere near Orlando.

After a year looking for work/post-docs (after a PhD in *i think* theoretical high energy particle physics) and no luck, Saad spent a few months learning MCAT material, took those, joined the army, went through OCS then into their medical program. He's almost done with his MD, and liking the field, the atmosphere, the fact that he got loads of his undergrad debt covered, and med school comped.

Umm, Josh (PhD in astrophysics sometime back around 2007 and not being able to find a job ... he was mainly looking for something in defense/government since he's a Bethesda native), got a job as a teller at a bank in Baltimore about 6 months post-PhD and has very quickly moved up over the years. I think they made him shift supervisor after 2-3 months, then assistant manager within the first year. He's now some regional loan manager / pseudo-quant. I think he's liking it and doesn't really mind that he had to start at "the bottom" right after he was done with his doctorate.

just some random stories of 3 of my friends/acquaintances and how they've dealt with not finding jobs in their fields post-PhD. Good luck though with getting something in your field, you still have time, but there are definitely options.

 

[twofish-quant]

I'm not sure how universities work in the states, maybe the HR filter is more "up front" than in the UK. But, surely, professors looking to (say) program PDEs in C++ would stress to HR departments that numerical Fortran programmers are acceptable!

HR departments are extremely busy and they forget. Typically, you give someone with no-science experience a set of keywords, and if they see the keyword on the resume, then they pass it on.

It's also easier to work it the other way. If you have numerical skills in Fortran, spend a month programming some basic C++ and you can add that to your resume.

The other thing is that it's assumed (and usually assumed correctly) that if you can handle C++, you can handle anything.

"Working in a bar" doesn't look great on the CV - good for keeping people skills polished

You'd be surprised. Something that I have seen happen is that someone with something unusual on their resume gets an interview because of that unusual job.

 

[twofish-quant]

There is some evidence to suggest that OOP Fortran is more efficient than C++.

Yup. Unfortunately unlike straight fortran, no one seems to have gotten OOP Fortran to interoperate with all of the other OOP languages out there.

There is some evidence to suggest that OOP Fortran is more efficient than C++.

I wouldn't be surprised if Fortran objects were more efficient than C++ objects, but I'd be extremely surprised if this applies to templated C++. The thing about templates is that you can get rid of the pointer call which allows you to vectorize loops.

 

[chill_factor]

Just a question:

I noticed that most of the job questions here are about hardcore theoretical physics, and how they should get jobs involving programming and finance.

However, there's a lot of physicists who do experimental work in things like condensed matter, who do not have a rigorous theoretical background and instead use commercial instruments to make measurements and the extent of programming required is Excel and maybe a bit of Mathematica.

How should experimentalists sell their skills, especially if its in a non-semiconductor materials field (biologicals, polymers, superconductors)?

 

[Locrian]

Just look at the things the company you're applying to needs. Characterization and testing are common ones - what type have you done and can do? Are you comfortable using/maintaining SEM/AFM etc?

Look at what projects you've started and completed and make them applicable to the work the employer does.

 

[elkement]

Just a question:

However, there's a lot of physicists who do experimental work in things like condensed matter, who do not have a rigorous theoretical background and instead use commercial instruments to make measurements and the extent of programming required is Excel and maybe a bit of Mathematica.

How should experimentalists sell their skills, especially if its in a non-semiconductor materials field (biologicals, polymers, superconductors)?

I was once in the position you have described - I went from superconducting thin films (prototypes for microwave applications) to steel (samples from production). None of the characterization methods or software tools was exactly the same.

I used the following selling approach that worked out fine:

I emphasized the general approach in characterizing high-tech materials, that is: how to organize standardized measurements (you could call this "quality management"), consolidate results gained from different types of measurements (electrical, microstructure...) and use an efficient approach to compare the results with production parameters.
Efficiency is key - I believe that in industry you have to prove that you are capable to use tools most efficiently without "re-inventing the wheel" (programming or developing stuff from scratch this is already available - because tools are likely to be cheaper than you labour costs).

There is a lot of menial and organizational stuff involved, such as setting up the measurement process, creating sample forms in paper or in digital form, involving the lab technicians, thus motivating other people... You need to prove that you are a hands-on guy and not an "absent-minded professor who wants to do deal with real research only" (This is a bias you can sometimes find in industry and I personally tend to say it is not unjustified sometimes.) I was "forced" to do also project management and controlling at the university -finally this help a lot to underpin my "down-to-earth / real live" approach.

In addition I pointed out the similarities in measurement techniques, such as TEM vs. SEM, X-Ray diffraction versus TEM dark field, sample preparation using ion mills vs. sample preparation using laser ablation etc. I sold myself as a physicist with diverse experience in different measurement techniques that would allow me to use any related technique.