Roman Kyslyi

I'm a data scientist with 4 years of experience in data science and 7 years of experience in software development in total including 3 years on leading and managing positions.

I have worked in different fields of data science, such as recommendation systems, NLP, time series. As a software developer, I have experience working with big enterprise solutions for telecom companies. During my career, I had an experience of successfully providing and implementing data science solutions for customer needs with leading and managing delivery team. Colleagues know me as a highly opened and reliable person who can always be trusted to come up with a new approaches and ideas.

I know that the client’s business comes first, and I never try to impose my ideas on others. Instead, I spend a lot of time understanding the business targets and expectations before suggesting ideas. I can work well alone, but I’m at my best collaborating with others. I have Masters degree in CS from Kyiv Polytechnical Institute, and I always try to be up-to-date and continue to study.

Data Scientist / Data Architect / Manager
email: [email protected]

phone: +380973589692

Summary

7 years of professional experience in the IT industry
Experience in Java / Python and related technologies
Experience in cloud computing (AWS, MS Azure) and BigData (Databricks, Spark)
Experience in Machine learning and deep learning
Strong problem solving skills
Strong organizational and communication skills
Interested in researching and applying my skills for problem solving

Experience

Deloitte & Toushe, June 2016 - Present

Data Scientist / Data Architect

Automation of Transfer pricing Benchmarking study

Responsibilities: leading development and data science teams, architecture design, building NLP models (english, ukrainian and russian), classification models, web development, information gathering, managing delivery team of 10 people.

Technologies: python, Flask, Tensorflow, scikit-learn, Azure, Kubernetes

Delivered scalable web solution with ML that currently is used in Deloitte globally for transfer pricing benchmarking automation (the tool reduces time spent by consultant on benchmarking by 40%).

Due to flexible architecture, functional parts of the platform are reused in other projects.

Platform for Merger and Acquisition

Responsibilities: architecture design, building time-series models (mostly working with company financial data), information gathering, managing data science team of 3 people

Technologies: python, Databricks (spark), scikit-learn, Azure

Digital maintenance tool (project kick-off)

Responsibilities: back-end development for MVP, architecture design

Technologies: Java 8, AWS

Statflo (part time, remote), July 2016 - April 2018

Data Scientist

Project: CRM system for wireless carriers

Responsibilities: prediction modeling, AWS Redshift migration, building NLP models, classification models Technologies: python (Flask, sklearn, gensim, Tensorflow), AWS (Lambda, Redshift, Glue), ETL (Pentacho)

Recommendations provided by developed models increased revenue by 13% for user on average.

U4G LLC, June 2015 - June 2016

Data Scientist

Project: U4G (social network for good things)

Responsibilities: recommendation system development, BI development, DWH development

Technologies: Python, ETL (Talend), architecture planning, managing development team of 3 people

Developed recommendation system for social platform, that helped to increase user retention rate

Software engineer, NetCracker, Oct. 2013 - June 2015

Software engineer

Responsibilities: product development, product integration

Technologies: Java, Maven

Samsung R&D Ukraine, Sept. 2012 - Oct. 2013

Software engineer

Project: Java obfuscator

Responsibilities: algorithm analysis, algorithms implementation

Technologies: Java SE, Android, Gradle

SAP AG Ukraine, June 2012 – August 2012

Intern

Project: Migration of SAP ERP on new DB

Responsibilities: configuration of SAP HANA DB, migration of data model from existing DB on SAP HANA, adaptation of SAP for using new data model

Technologies: SAP R/3, Oracle 10g.R2, HANA database, SQL, ABAP4

Technical Skills

Programming languages:

Java, SQL, Python

Databases:

Oracle, SAP HANA, PostgreSQL, MySQL, MongoDB, AWS Redshift, Redis, Elastic Search

ETL:

Talend, AWS Papeline, Pentacho, Airflow

Machine learning:

Python, Numpy, scikit-learn, pandas, spacy, NLTK,
Tensorflow, Pytorch
traditional machine learning algorithms (classification, regression, etc.), regularization techniques
Experience building neural networks with different architectures - FCN, RNN, CNN
Knowledge NLP techniques like topic modeling, language models, sequence to sequence models, attention
traditional NLP techniques (count based distributed representation, part of speech tagging, dependency parsing, Hidden Markov models, CRF, LDA, etc.)

Big data tools:

Hadoop, MapReduce, Spark, Databricks

Frameworks

Spring and Spring Boot, JPA Hibernate, Java concurrency (Executors, Fork/Join), Flask, Celery

Deployment and CI tools

Docker, Docker Compose, Ansible, Terraform, Zappa, Jenkins, Circle CI, Kubernetes

Education & Certification

Institute of Applied System Analysis. Kyiv Polytechnical Institute, 2009 - 2015

Bachelor of Computer Science

Master of Information technologies in system design

Udacity, Feb 2017 - Dec 2017

Self-driving car nanodegree

Stanford online, Feb 2015 - Feb 2015

Statistical Learning cource

Papers

DETECTION OF HUMAN RESPIRATION PATTERNS USING DEEP CONVOLUTION NEURAL NETWORKS (http://journals.uran.ua/eejet/article/view/139997)

Additional information

Languages

Ukrainian - native
English - advanced
German - intermediate
Russian - advanced
Chinese - elementary

Additional skills

Experience of leadership of sport groups (up to 15 people)
Experience of organization public events