Hi,
I'm Sergey
Data Engineer

Keen on Data / Analytics

You have probably...

...got in here because you are holding my CV. This web page
is just a short description of my life/work and also serve as
a supplement to my CV.
I worked for a couple of different companies helping them to
orchestrate their data and to derive value from it. What responsibilities and tasks were on my job is written on CV. How I got into data engineering/science? Well...

Well...

... it was not easy like many other things in our life.
You constantly question everything in order to understand how something works. Sometimes, you got it immediatedly, other times you struggle, search, ask, search and then you got it!
However, when you look back you think it actually was not so difficult. Which means that you gained some experience.

My path started as Linux Admin and now I'm Data Engineer or in other words I'm in the world of big data. It is just a hype word for a huge amount of data as well as Machine Learning which is just Statistics and some Math.

I began learning about data as data can have many flavours, shapes and types. There are whole lots of about data like how it is stored and how the way of storage can help you to solve you tasks. When you have to deal with massive data you have to work with alternative tools such as Hadoop, MPP and Spark to name a few, as using ordinary tools is not always effective. So, you established your data lake, figured out how to collect you data and clean it. What comes next? The next step is ...

Next step...

... is to provide data to consumers/other departments. Phew... the job is done.
Not really, as a data engineer you also closely work with Data Scientist/Analysts. This is where you apply your knowledge of Stats and Math (if you have one) 🥸. I think this stage can be devided into two parts.

Part I. You try to discover what data is trying to tell you. This is very important skill which I think only a few people have. Then you usually plot the data, analyze, pick the most significant piece of data and build visual dashbords.

Part II. Once the data is in right format you sometimes need to do feature engineering. Find out what features are important, which is called correlation, to predict certain labels. You usually try different ML algorithms based on the task which can be classification,regression, clustering etc. to find the best result by looking at metrics.

I have applied XGBoost, CatBoost and scikit-learn. So, I will tell you that I really like them because they are pretty easy and there is a lot of infomation available on the web.

The end

That was just a short story about myself. It might help you to get a small picture about me.
So, drop me a message by filling up the form below or send an email to t.sergey-94@yandex.com

Developed in 2016

>