I currently live in the Dallas-Ft. Worth area and work as a Data Scientist for Capital One. We are creating software to make model rebuild work-flows when the models have complex interdependencies. We will use this to rapidly iterate on model rebuilds and experiment with methodologies. I get to play with the cool Big Data tools.
Before Capital One I worked as a process engineer and then research scientist in the semiconductor industry for 4 years. I was responsible for shooting high-energy Helium and Hydrogen ions into silicon wafers so that we could peel off really flat, really, really thin layers and leave them behind on other wafers. How would you cut and paste a layer of silicon atoms that was 300mm in diameter and only 1000 atoms thick? Anyway, it resulted in a Silicon + Oxide + thin Silicon sandwich (called SOI) which was then probably used in your iPhone’s antenna circuitry. My research was around how to do this better and why things no one understood happened anyway.
Before that I was studying physics in school and working on projects with nice people in Missouri. One project was to make nano-tubes so that we could melt LiBH4 into it and then see what happened when we changed its temperature since it would be so small. We wanted to do this because if you want to get the hydrogen out of LiBH4, perhaps to run your H-powered car, you would have to make it very hot, but that wasn’t good because that takes energy. It turned out that when you make it really small it gives you H at a much lower temperature.
I also learned to use a Transmission Electron Microscope (TEM) to look at all the small things we were making. This also involved image analysis and simulation work. For instance, in a different project I looked at pixel intensities over single, heavy atoms in TEM images of star dust to try to figure out what kind of atom it was and therefore what kind of world the star dust might have came from.
Before college I mainly played guitar, which I do now with less frequency. Now I spend more time programming and working on interesting projects. At the time of this writing I’m building algorithms that generate fantasy baseball rosters. I’m currently working on Markov Chain techniques to generate outcome probability distributions. I’ve also customized the Ant Colony Optimization scheme for solving the constrained knapsack problem (i.e. pick my baseball players).
Next, I want to build an IoT device that incorporates online Bayesian learning. I’m still considering the details. Find me on LinkedIn to chat. Oh, and I really like my two dogs.