How I left academia

From stacky wiki

<protect>Anton</protect>

Several people have recently asked me why I decided to leave academia, so I thought I'd write it up. The short version is that I tried to stay in academia, but wasn't able to.

The story begins my last year in grad school. While I was applying for jobs, I got a call from a headhunter who told me that he thought I'd be a good fit for a company called Virtu Financial. I told him that I wasn't really interested in a non-academic job, but that I was curious about it. He arranged for me to talk to a couple of people he'd placed with Virtu the previous year. I was glad to talk to them, but I was set on an academic job.

I knew my wife was coming to Caltech for a postdoc, so I applied to jobs in the LA area. (One of the reasons the job at Virtu didn't appeal to me is that it would be in New York.) I also applied to a quasi-academic job at CCR (Center for Communications Research), a division of the Institute for Defense Analyses which does math for the NSA. I wasn't allowed to know what math they did, but I'd heard from multiple sources that it was an excellent place for a mathematician to work. That description is completely in line on my interview experience there. For my personality, it's even better than an academic job since it is highly collaborative. Luckily, the feeling was mutual: I ultimately got an offer from them.

It was especially lucky because I didn't get any other offers, which surprised me a bit. I'm really proud of the research I'd done up to then, but I didn't have any publications (at the time of this writing, I still don't). I realize that on paper I didn't look spectacular, with the exception of founding MathOverflow, so I arranged to visit the area in early February. The idea was that if I invite myself to speak in various seminars, it'd be easier for people to see that I'm a good mathematician. That plan didn't work. I was only able to arrange to speak in two math departments. At one of them I learned that the position I'd applied for was promised to another field, so I had no chance. I later learned that the other one couldn't make me an offer for similar reasons, even though they thought I'd be a great fit.

I happily accepted the offer from CCR, but in late April I was told that I didn't get security clearance, which was required for the job. I wasn't given an official reason, but it was pretty clear that the problem was that too many of my immediate relatives are not US citizens, and that every such connection increased the amount of money it would take to investigate me in order to grant clearance.

One of the MO moderators had the idea that it may be possible to apply for money for me to do some work related to MathOverflow, and that I could arrange for an academic host in the LA area. That way I'd still be able to do mathematics, I could apply for academic positions again the following year, and I'd have a departmental affiliation in the mean time. I contacted the Caltech and Harvey Mudd departments, which were both very supportive. Ultimately, the funding didn't materialize, but miraculously (to me) Tom Graber and the Caltech department were able to get together enough money to offer me a 1-year position.

I'm extremely grateful for that year. Even though I wasn't able to extend my stay at Caltech (I hoped to get the NSF postdoc fellowship, but didn't), and I didn't continue in academia afterwards, I got to do some really excellent research while I was there. I also got to do some really excellent teaching too; every term, I got to teach one course I'd wanted to teach at some point in my mathematical career.

The next year I cast a much wider net and applied to academic jobs all over the country. It didn't work; I didn't get any offers this time around either. During the process, I heard again from the headhunter for Virtu. This time I was more receptive, particularly since there was now an office in Santa Monica, which is pretty close to Pasadena. I also chatted with Michael Kleber, who was an (awesome) associate professor at Brandeis when I was an undergrad there. Michael has worked for Google for the last several years, and encouraged me to apply. I ultimately got offers from both Virtu and Google. Both jobs involved working in a great environment with very smart people on interesting problems whose solutions make the world a better place, so I couldn't go wrong.