Benjamin Gaska

I'm a Software Engineer who currently works at Google on the Drive Team in Boulder, Colorado. My work at Google has been on the Drive Data Protection Team and has focused on controlling Googler access to end-user's data. This has largely focused on the creation of infrastructure that enforces proper access pathways for users. It also includes monitoring, logging, and enforcement of regulatory requirements. Specifically I've built several frameworks for specifying access controls, and monitoring for regression in systems all across Drive. This has been adopted across Drive and has been selected as the official means of access control in the organization. My work has involved working with all teams in Drive Backend, as well as broader work with Docs, G-Mail, and security teams in general. Before that at Microsoft I work at Bing Local - Premium Data. This team was focused on using data available to us to improve user experience and presenting more useful information immediately to the user. That always sounds vague to me, but I think my specific tasks make it more clear. In the year I was there I worked on two significant projects. First, I did the E2E restaurant review pipeline. This used ML models to extract relevant phrases from restaurant reviews, organize them into different categories (food quality, ambiance, parking, cleanliness, etc), and presenting it to users along with a histogram of reviews and other, hopefully immediately useful, information. I was continuing work on this, as the model was still in a simple place when I left, but this can still be seen by searching for a restaurant in Bing. Second, I was working on the Bing News application. I was about 2 months into this project when I left, and I had been doing basic data extraction pipelines, once again focused on Restaurants and Tourist destinations. Before that I did research in NLP and Distributed Systems at University of Arizona. Most of my work involved large dataset analysis or programming language work. I still have a significant soft spot for API and Language design, along with NLP.