Alumni Early Career Profiles - Keith Schon
|Education:||B.S., Mathematics and Ecology and Evolutionary Biology, The University of Arizona, 1997
M.S., Mathematics, Stanford University, 1999
|Position:||Sr. Software Engineer
Since late 2004, I've been working at a start-up company called Cataphora. We analyze the unstructured electronic data that all of us create on a daily basis (email, IMs, system logons, etc.) to reconstruct patterns of behavior, both for individuals and for organizations. Most of our business involves civil or criminal litigation, or internal investigations in large organizations.
The amount of information generated (and stored) by the average US worker is growing exponentially, and the amount of information a single person can read and understand is not. This means that applications which require organizing and understanding all that data — for example, responding to a subpoena — are constantly getting more expensive. I write software to help with this. We provide fine-grained categorization of data, analyze organizational structure in terms of command structure and information flow, detect normal process workflows and deviations from them, and detect attempts to delete evidence. In some cases, we detect attempts to avoid leaving evidence in the first place.
I received my B.S. from the University of Arizona in 1997 with majors in Mathematics and in Ecology and Evolutionary Biology. I also received a M.S. in Mathematics from Stanford in 1999. My work involves graph theory, abstract algebra (lattice theory and linear algebra), probability, and statistics. Many of my coworkers have backgrounds in computational linguistics and social network theory as well.
It's fascinating to work on the leading edge of what is possible. This is especially true because we are in a field that affects so many people. The ethical and logistical aspects of this sort of data analysis are still being worked out by society, and I have a chance to affect how these technologies are used.
My advice to a math student thinking about his/her future career is to look for subjects you're excited about. Some classes you take will provide skills that are applicable in all sorts of different environments, for instance statistics and computer programming. You're less likely to use some other mathematical subjects directly (although you never know), but if you're excited about them you'll do interesting things with them. A lot of employers are looking for candidates who have learned strong mathematical problem solving skills and have shown that they can do original work. That can be worth more than having a specific skill set, depending on the situation.