Recently I have been working on a music recommendation system and audio and video fingerprinting algorithms. Before that I spent a lot of time working on embedded systems (mostly Linux-based). I have a good understanding of how computers and operating systems work and I love applying it to build super-efficient software.
Software Engineer @ From January 2015 to Present (1 year) Backend Engineer @ Together with a colleague we created the first version of the continuous play sound recommender for soundcloud.com. Later we designed and implemented a distributed version of this system that can handle the growing amounts of data produced by the site. From November 2012 to August 2014 (1 year 10 months) Software Engineer @ My main goal was to improve indexing of audio, video and melody fingerprints at large database sizes. This included both designing new algorithms and tuning existing ones. I have also participated in running precision evaluations using human raters. From October 2011 to November 2012 (1 year 2 months) Software Engineer @ I worked on various projects involving embedded devices running Linux. My work consisted mostly of writing device drivers for different devices. Also I worked on several frameworks that decouples video and image processing algorithms from the actual hardware they run on. From 2006 to 2009 (3 years) Software Intern @ Developed a mechanish for obtaining driver crash dumps and tools to post-mortem analysis. From June 2008 to August 2008 (3 months) San Francisco Bay Area
M.Sc., Computer Science with focus in Visual Computing @ Eidgenössische Technische Hochschule Zürich From 2009 to 2011 B.S., Computer Science @ Sofia University St. Kliment Ohridski From 2005 to 2009 High School Diploma, Mathematics and English @ Sofiiska Matematicheska Gimnazia From 1997 to 2005 Boris Daskalov is skilled in: Algorithms, Data Structures, C++, Systems Programming, Distributed Systems, Music Information Retrieval, Computer Vision, Machine Learning, C, Linux, MapReduce, Software Development, Subversion, Embedded Systems, Pattern Recognition