I am Niko, a Software Engineer and Architect, an Applied Mathematics enthusiast and an all-around hobbyist. My main interests are at horizontally scalable architectures and algorithms, data science and machine learning and some less known programming languages such as Clojure and CUDA. Professionally I don't mind getting my hands dirty with C++ when high performance is required but usually I find myself writing Matlab, Python or Bash scripts. I try to keep my up-to-date with all of these technologies but of course there are too many stay 100% on track on everything that is going on. Technology and frameworks are evolving so fast! This picture is a summary of technologies I have used so far on projects:
I have always done small projects during my free time, but before this blog I just had folders laying on my computers and maybe some screenshots in Dropbox. It was difficult to share them with others and in retrospect remember which major projects I had done in recent years. This motivated me to start compiling this blog, but first I had to write my own blogging engine ;). It is based on Git for SCM and CMS, LaTeX for writing, PHP to generate navigation links etc. and mustache templating to generate the final HTML. More detailed info can be found from the Blogging platform article, but actually now that I'm past PHP stage of my career I'm in the middle of a long process of porting this engine to Clojure. Currently static output files are hosted at Github pages and optional PDF files are in Dropbox's public folder (HTTP 302 redirected from a micro AWS instance at cdn.nikonyrh.org).
As a brief note about my career, I joined Analyse2's Analytics Team in April 2014 where we (a few doctors and masters) focus on predictive modeling of product sales volumes of hundreds of thousands of products in thousands of stores. It has been a great opportunity to collaborate with other smart people, build PoC infrastructure and see it grow from internal test environment to production, introducing new tech to make our lives easier and scaling peak computational needs to Amazon Web Services with PySpark.
Between 2010 and 2013 I spent two years in Singapore: had a student exchange program at the National University of Singapore, and an internship and full time position at PropertyGuru. Initially my main role was a PHP & MySQL developer, but later it evolved to data analysis in a newly formed team. The main focus was on an automated property valuation algorithm and to make it available at RESTful API. It might still be publicly visible at api.propertyguru.com (/v1/properties/[id]/valuation/...), but naturally queries aren't accepted without valid OAuth authentication.
I hope to keep this blog reasonably active, but not all my projects are worth the effort of actually writing an article out of them. Some are just small steps towards a greater long-term goal.