Декабря 2015 - по настоящее время
Freelance work along fulltime job - mainly a lasting text summarization project for English classical literature I'm developing on my own for more than a year now.
Needed to get outside Yandex experience a bit - learn NLTK, scikit-learn, for instance. TensorFlow is next. And to get experience of faster development with no production to prevent from crashing
Researcher & Developer
Июля 2012 - Июня 2017
5 years in NLP. First 3 years developing a Named Entity Recognition system for queries and improving its quality (more as a dirty-handed researcher writing scripts and classifiers), later on refactoring and rewriting a big system generating query expansions from various sources (getting less creative but finally a decent coder).
C++ for production, Python for scripts, internal Yandex libraries for any task (even special internal string container instead of stl::string). Local internal version of MapReduce for storing data, internal ML libraries, internal product to represent computations.
Quite the coder you expect from a big company - patient to trying not to crash the production and very company -specifically trained.
Июля 2011 - Июля 2012
Interning in the Linguistics department along last year in university - the NER system extracting and classifying entities I later on developed when hired fulltime, learning to code, handle big data (and search data in general), completing university diploma (and basing it on the work I've done for production)
Intellectual Systems for the Humanities
Russian State University for the Humanities
2007 - 2012
(5 лет) Moskva