Driver Org: Building Uber’s platform for flexible work.

Search Relevance Team: text tokenization, query parsing and reformulation, data analysis, logging instrumentation, metrics, machine learned ranking, lucene/solr.

Prototyped a query suggestion model to compute semantic similarities between Knowledge Base (customer support help articles) queries mined from the search logs. The prototype was fully automated to run daily updates and expire old data and was built it with all the cool toys of the time: HDFS, HBase, Hadoop, Pig, Java MapReduce.

Worked all over the stack on Cisco Quad, an enterprise social networking platform, from front-end features including message inbox, reporting structure, and profile page to backend features including social activies architecture/permissions and search boosting/facets/indexing. Worked with: Lucene, Solr, XMPP, JMS, AMQ, Oracle RDBMS, MongoDB, Dojo, JQuery.

As a senior research project in software engineering, I worked on UrbanSum (used for integrated planning and analysis of urban development) and replaced the map drawing libraries with Mapnik and integrating GIS software tools and data formats. The result was that all the bitmap raster graphics maps were replaced with configurable and improved, smoother vector graphics maps.

Interned with Boeing's International Space Station (ISS) Payloads Software team contracted by NASA and updated some software tools that generated binary config files for payloads equipment sent to the space station in ISS Expedition 18. Also worked on test software used in the equipment lab. This was at the NASA/Army Marshall Space Flight Center where I got to explore space station hangers and battle simulation rooms, and even meet astronauts!

Worked on tools for comparing DBs holding all 787 Dreamliner data. Also spent a lot of time playing with software that could draw a 3D model of the 787 Dreamliner with only the plane parts that you select from those DBs. It was CAD for plane engineers and it was awesome.