CERN/SLAC : INSPIRElive site
I spent 4 months in Geneva and 8 months in California working on the INSPIRE project, a joint effor between CERN and SLAC. The basic story is that high-energy physicists have a much more open academic community than many other fields, and they publish much of their work in the form of pre-prints on arXiv. However, arXiv is intended as essentially an RSS feed of new papers, while searching functions on those papers have long been fulfilled by another service called SPIRES. SPIRES was turned off by SLAC and has been replaced by INSPIRE, for which I built some query parsers (mainly to support legacy SPIRES syntax), some nice UI tools (search indexes on figure captions, and the ability to see a paper's figures without downloading the whole paper), and some backend optimizations.
Blue Oxen Associates : Analysis and WikisI worked with Eugene Eric Kim for several months on a project for a Fortune 100 company. I did social analysis of email logs and built a single sign-on plugin for MediaWiki.
Nuance : Secret!I built a secret tool for Nuance over a half-year which I can't share the details of at this time. I can say, however, that it was a web-based tool for a database of audio data which included converters and other handy things. I designed the interface and implemented it top-to-bottom myself.
Indiana University : Undergrad InstructorEvery semester that I was at IU, I was involved in some kind of teaching exercises, and every semester it was for a different course. In my first semester, I was a peer tutor for the intro CS class, taught in scheme. From there, I moved on to undergrad instructor for the second course, taught in Java. Data structures was next, and the following semester it was computer structures. I finished up with design patterns, a graduate class.
Technische Universitaet Darmstadt : Guest ResearcherI spent a summer in Germany working on NLP projects: in particular, I was developing a new algorithm for back-of-the-book index term generation which, instead of crawling the whole book at once, first split the book into what it believed were "sections" and then performed analysis on those sections in order to get an even spread of terms.
Google : Knol (cum Docs)live site
Back in 2008, Google experimented with bringing personal credibility back to the internet and shifting away from the anonymity and untrustworthiness of Wikipedia. While this experiment ultimately failed, I did work on the text editor to support folks trying to share their knowledge wiki-style. The two projects I worked on during my summer there were the special character insertion dialog (before I touched it, the only option for insertion was typing in a unicode code) and the LaTeX inserter/editor.
Indiana University : Undergrad Researcherlive site
My first summer in school, I worked as an undergrad researcher for Geoffrey Brown/Kam Woods on a project for accessing CDROM data that was built with long-unusable programs. I helped build everything from the interface to the backend conversion scripts. It has gone through numerous iterations since then, but the spirit is still there.