Working at Yelp, bouldering (on v6's now), taking care of Oliver, volunteering, occasionally reading, taking a coursera course on voting, taking short motorcycle trips, watching anime, sometimes socializing (I know, what?!?).
I'm a software developer at Yelp. I worked on the search-infrastructure team integrating ElasticSearch into our production environment (I probably shouldn't say much more publicly), and am now working more on the quality side, on a team aiming to curate Yelp's datasets.
I have scripting, design, and communication skills. I am capable of both hacking and writing production-grade code. Most of my experience is with mid-layer server software, but I am generally competent in Linux environments, and have written a bit of frontend code (d3, Angular, REST, selenium) for internal sites.
I also have some experience with conflict resolution—I've started conversations like, “Hey, I'm really sorry when I said X; it came off wrong”, or “I didn't entirely like how you brought up X”. Personally, it's challenging, and something I've had to learn. For the record, all of these discussions at Yelp have turned out well, and led to a more positive work experience.
I have a lot of interest in research, but haven't produced anything groundbreaking yet. Possibly related, I'm somewhat transitioning from programming languages work I did in my undergraduate years (as a research assistant), to machine learning (I'm particularly interested in neural nets and NLP). Here's my cv, but it doesn't usually get updated that often.
Also, I'm sorry to disappoint, but I'm not currently [as of Sept 2014] available for hire.
I was a student at UC Berkeley, under the supervision of Ras Bodik. I worked on tools to facilitate development of GPU programs, using SKETCH and Haskell EDSLs. A while before that, I independently built a CUDA video coder using GA-tuned lifting wavelets. (More details on my CV).
I wrote a small open-source fork-based parallelism library vimap, which was developed partly at Yelp. Vimap is primarily aimed at small tasks, like ad-hoc analysis or data loading [e.g. into ElasticSearch], when you don't need the scale of map-reduce or the like, but serial execution is too slow.
As of Sept 2014, I'm working on a very minimal library plcd to facilitate use of pre-loaded compression dictionaries.
Here's my digital footprint, if you want to do some stalking (just don't be creepy): g+ page (personal); flickr (road trips + stuff); blog (tech notes / links); stackoverflow (programming q&a site); github (code samples, etc.); bitbucket (code samples, etc.).
I'm pretty out as genderqueer and physically transitioning / transsexual-like. The latter term has various gender identity implications which don't hold true for me, so I usually just call myself "genderqueer" if/when necessary. I use gender-neutral pronouns and language (I don't care which neutral pronoun, just try to gender me correctly; I suggest ze/zir or they/them/their). [ More on pronouns ]
Lyon-Martin, the ACLU, Democracy Now, Amnesty International, Breast Cancer Action, the Ada initiative, Planned Parenthood. Things I no longer financially contribute to / fully support: the HRC with its hetero-assimilationism.
I have a few random subpages; they're very casual writing and I'd appreciate you reading them generously (don't nitpick, try not to be offended if i wrote something stupid). [ quick thoughts on invasive body questions ] [ body butter ] [ gpu image processing setup ] [ my vimap notes ]
email: [gatoatigrado at gmail]; cell: 443-267-8864