How do Graduate Admissions Committees view an applicant from Industry?

Question

At the time of writing this, I have been working for Raytheon Missile Systems (RMS) for about 2.5 years since graduating college.

Prior to graduation, I was on track to go a more academic route due to graduate work and independent study projects I was involved in related to computational mathematics. I ended up choosing to try out industry first, however, since I had no experience with it and wanted to know if this path would be preferable for me long term.

While I have been successful at RMS and used my computational math background to achieve a lot there, I don't think industry is where I want to stay and so I am now working on applications so I can pursue a PhD and more academic career.

Given my industry background, are there any common beliefs in academia, whether good or bad, about individuals from industry I should be aware of for the sake of my application? Is there anything I should strive to highlight or avoid in my application that would help me fare better in the application process?

EDIT: My experience pre-RMS was largely in numerical analysis, optimization, and simulation, with emphasis in nonlinear systems of hyperbolic partial differential equations. One sample project, for example, was implementing a Space-Time Discontinuous Galerkin Finite Element code to tackle a system of Hyperbolic PDEs and doing validation and convergence studies to validate the theory. I also gained decent experience via independent studies and working at NASA JPL in computer vision. Other skills gained were in parallel computing using OpenMP and MPI.

At RMS, I have worked on clusters and have primarily developed and built codes in areas like numerical optimization, adaptive sampling of unknown functions, supervised machine learning, controls, bayesian estimation algorithms, and simulations.

In terms of research, I am planning to apply to computational math or computer science programs and hope to get involved in machine learning with emphasis in Reinforcement Learning for partially-observable Markov decision problems in continuous state and action spaces.

Answer 1

There are several considerations that may make it difficult for you. Here are a few based on gut instinct:

1) You are an 'outsider' - the traditional route of undergraduate to graduate school often entails nurturing relationships with other academics at your home institution and abroad. Those relationships help grease the wheels when it comes to letters of support. But, this can be overcome. Perhaps you already maintain relationships with academic circles. Failing that, strong letters from higher-ups at a place like RMS are probably just as noteworthy.

2) From the PI's perspective, taking a graduate student is a long-term investment. You would occupy a spot that could be offered to someone else, so in addition to promising productivity, PI's generally prefer the person have the goal of continuing on and pursuing an academic life. Since you've been in industry, there is always the fear that at some point, you'll simply abdicate and return to industry.

I think you can spin your industry experience 'narrative' to your advantage. First, you can stress that your time in industry did not meet your needs for academic development. In other words, only academics provides the platform to realize your highly technical goals. Second, people in industry usually have a very strong work ethic and know how to get things done; something that is often lacking in new graduate students. You might consider documenting just how productive you have been.