We’re currently building new back-end identity and user-profile services that can be used on chat networks. Identity services are mostly for authorization, where the users have to prove who they say they are using a thumbprint or password. Profile services contain information about user preferences, purchase history, favorite items, Facebook likes, and more. A lot of the things we build are designed to appear where Millennials spend most of their online time, namely messaging. Instead of asking people to download an app or visit our website, we just “show up” and let them find great eBay deals on their favorite messaging platforms, including Skype, Facebook, Snapchat, Instagram, and others.

A big breakthrough occurred when our team developed the “bad-buying experience” prediction model. The model enables eBay to demote the search relevance and completely filter out items in the eBay inventory most likely to give our customers a bad experience, such as items out of stock, not as described, or shipped late. We also recently discovered how to measure “silent suffering,” where users do not report a bad experience they have had. The discovery has had significant impact on behavioral economics and the definition of transaction defects.

A lot. My favorite thing is the people I work with. Our team of collaborative, excellent engineers works tirelessly to build great customer experiences. It’s also a fantastic work environment here at eBay. There’s free food, a gym, and my office is close enough so I can walk to and from work.

I started out as a manager, but now I’m a director. I have moved between engineering and research a few times, and have gained a lot of perspective from being involved in the theoretical and operational sides of the business. Next, I hope to take on more DevOps responsibilities to get the public cloud systems we use more integrated into the core of eBay.?