I am a common mathematician, thoroughly mathematical by training, temperament, and trade. I’ve trod weary footpaths through numerous industrial sectors, plying my skills in areas as diverse as banking, transportation, telecommunications, retail, and even venture capital. My appetite for mathematical application is quite ecumenical; although I focus on optimization and discrete mathematics, I embrace any opportunity to contribute to the world amenity through judicious employment of mathematical principles and practices. Yes, I am a common mathematician.
It was not always this way. My original intent upon entering university was to become a biochemist. Like many scientists before and after me, I aimed (in no particular order) to discover a cure for cancer; find a solution to Type 1 diabetes; win a Nobel Prize in Chemistry; and not set fire to, blow up, or otherwise damage any laboratories or chemical facilities. As luck would have it, I achieved none of these goals. Indeed, my migration from chemistry to mathematics was precipitated by an unfortunate “incident” in the laboratory of my then-advisor, James Collman, when a combination of my maladroitness and ignorance inadvertently caused a reaction vessel to catch fire. The details, I believe, are unnecessary [ahem].
Suitably abashed, I abandoned my goal of becoming a chemist of any stripe. Instead, I accepted a lateral “promotion” within Stanford University’s School of Humanities and Sciences and switched my major to mathematics. In those days (the late 1970s and early 1980s), one could receive either a B.A. in (pure) mathematics or a B.S. in (applied) mathematical sciences. Being of a pragmatic bent, I chose the latter, which sparked my interest in operations research. Excess credits from both advanced placement classes in high school and an aggressive freshman course load afforded me a chance to simultaneously pursue my bachelor’s and master’s degrees. I leapt at the opportunity to save time and money while scoring two degrees within four years at an expensive school like Stanford.
Deep thoughts, deep learning — it is all of a piece for Kevin Gillette, a common mathematician at Accenture Federal Services. Photo credit: Joon Yoon.
Upon graduating with both degrees, I faced the perennial question that many young SIAM members encounter: What comes next? I had always fancied being a college professor, and contemplated working towards a Ph.D. in math or a related discipline. But just how much PTA (pain/torture/agony) would I endure along the way, between the whims and vicissitudes of advisors and faculty and the subsequent POP (publish or perish) mentality that pervades academia? Achieving tenure requires suffering for one’s art, and I’ve never savored the idea of being a starving artist, irrespective of the art form (yes, mathematics is indeed an art form).
My practical and productive upbringing steered me towards an industrial career. I had happily completed an undergraduate internship at Bank of America and evidently left a good impression on the team; they invited me to join them full-time upon graduation, which I did. Mathematical finance is now all the rage, and ample job opportunities—and commensurate salaries—exist for mathematicians who enjoy working with Ito calculus and stochastic differential equations. In the early 1980s, when such techniques were still relatively inchoate, the practicing mathematician had two basic paths: actuarial work or general consulting within the company. My team’s title, “Management Sciences,” aptly described its role. Our projects on behalf of the bank included issues like float management (cash that the bank can use for investments before it posts against account ledgers), portfolio diversification, and risk scoring. I spent 18 months working on these types of problems. My marriage and a desire for different work caused me and my new bride to relocate.
I moved to Dallas, Texas, where I continue to reside. My second industrial setting was transportation — specifically American Airlines, headquartered in the Dallas/Fort Worth area. I was invited to join their Operations Research department. Interesting work abounded there, including network optimization problems (crew assignments to flights, repair of crew and passenger assignments during weather complications, etc.); inventory-theoretic problems (rotable parts inventory stockpiles); queueing-theoretic problems (gate assignments in real time); and simulations (which replicated how complicated airports like Dallas/Fort Worth or Chicago O’Hare operated under different gating protocols, call center resource distribution, and so forth). I also applied my tradecraft as a systems support analyst in the Operations Engineering group. This team comprised the lion’s share of American Airlines’ aeronautical engineers and performed mission analyses for both existing and prospective aircraft fleet types, flight planning system studies and data grooming, and weight-and-balance examination for individual flights. I became an in-house expert on map projection equations and techniques for calculating route distances and headings — fairly quotidian calculations, but vitally important to get right the first time.
Since then, I have served in a wide variety of positions and contexts. I worked for nearly five years as a de facto senior engineer with MCI, where statistical analysis was exceedingly important. Following my tenure, I engaged in numerous contract programming and analytical assignments. One of them was a small (two-person) venture capital experiment where I vetted hard-science investment opportunities (chemistry, physics, geophysics, life sciences, etc.) for a select list of investors. This job required that I read through and scour nearly 50 peer-reviewed journals each month to get a feel for current research.
I also spent four years working at Blockbuster LLC prior to its ultimate demise. The need for statistical knowledge once again became paramount, as our business was primarily devoted to product placement and assortment. My team met the challenge of demand forecasting with a variety of techniques, most of them standard (principal component analysis, clustering, autoregressive integrated moving average, and so on).
All of this experience served as prelude to my current position as a principal in the analytics practice at Accenture Federal Services, a domestic subsidiary of Accenture LLP. Although I am technically a senior data scientist, my work for the past eight years has covered everything from predictive analytics (forecasting and predictive modeling) and large simulations to exploratory data analysis on very large (1+ terabyte) datasets. This rather serpentine career path has yielded the following insights, which I will share primarily for the benefit of those who are just beginning their trek beyond the academic world:
1. Listen more than you speak. You already know that you are intelligent and learned; what you still need to learn is everything else. I say this with tongue only somewhat firmly in cheek.
2. As with any challenge, the greatest difficulty is knowing how, when, where, and to whom to ask the right questions. In both research and industry, it is much easier to address a well-posed rather than an ill-posed question. Moreover, whatever solution you devise will be that much easier to explain and defend since the correct question essentially becomes its own evaluative rubric.
3. Always work to increase and extend your command of English (or whatever your business language may be). Few things can derail an otherwise splendid work product, analysis, or result faster and more thoroughly than an inability to express its meaning to someone who matters.
4. Teamwork is essential, especially in the industrial sector. I can recall very few colleagues over my 36+ years in industry who have “gone it alone” and succeeded. Moreover, the work becomes much more rewarding with an excellent team (I speak from experience). It is great to share successes and commiserate failures with others.
5. Be generous but humble with your own knowledge and experience. I have had countless mentors over the years and am extremely grateful to all of them! I have also had the signal honor of mentoring a few people along the way. It’s quite a rush!
All of this, combined with the various quisquilia of industrial mathematics at large, prove my initial proposition: I am a common mathematician.
How about you? Are you doing interesting work, or do you have a unique career trajectory? Write to us at [email protected]! We may publish your account in an upcoming issue.
||Kevin Gillette is an analytics principal for Accenture Federal Services. He can be reached at [email protected].