Особисті дані приховані

Data Scientist (Machine Learning Engineer)
Oleksii T
Location Kyiv or Lviv, Ukraine

Phone number: [відкрити контакти](див. вище в блоці «контактна інформація»)

Telegram @ginger_snaps

Summary
Experienced Data Scientist and Quantitative Developer with 15+ years of expertise in statistical
modeling, machine learning, and algorithmic trading systems. Proven track record of designing and
delivering end-to-end quantitative platforms, time-series forecasting models, and large-scale
back-testing frameworks for financial markets. Strong background in data engineering, production
model deployment, and applied research across domains including NLP, DSP, Bayesian A/B testing,
and marketing analytics. Deep understanding of both classical statistical methods and modern
machine learning approaches, backed by a rigorous mathematical foundation and a hands-on
systems mindset. Recently focused on LLM fine-tuning (LoRA/QLoRA), RAG pipelines, and agentic
AI systems, integrating theoretical depth with practical ML service deployment.

Skills used in real projects
● Classical Classification and Regression including Cluster analysis and Machine Learning
● Bayesian approach especially for A/B testing
● Deep Learning approaches applied across the methods mentioned above
● Extensive experience in data wrangling, ETL, and statistical preparation. Any sources, any
environments (cloud or not), any databases.
● NLP, Time-series data, DSP
● LLMs & Generative AI: LoRA/QLoRA fine-tuning, RAG pipelines, HuggingFace Transformers
ecosystem, PEFT, agentic AI architectures
● Сreating complex sample designs, marketing research
● Good visualization and programming skills

Additional Information
● Python: DS-stack libraries + DL libraries (PyTorch/TF) + specific libraries like pymc3, pyod, etc
● Additional Python libraries for web (flask/fastapi), scraping (scrapy/selenium), message brokers,
parallel/asynchronous/distributed computing (asyncio, PySpark), cloud, etc
● R: used over 70 libraries in practice. Working with R since 2012.
● In some projects, Julia, Scala, SAS, Q/K languages were used as the main PL.
● SQL-like DBs, NoSQL DBs, column-oriented DBs.
● Rewriting algorithms to general-purpose PLs like C#/Java
● Bash, Docker, Spark (pyspark preferably), AWS (SageMaker, S3, RDS)
● SPSS Statistics (expert level in macros and “syntax” scripts)

Language skills

● English – upper intermediate
● Ukrainian – native
● Russian - proficiency
 Education
Taras Shevchenko National University of Kyiv (2009)
Faculty of Mechanics and Mathematics, Mathematical Modeling, MSc in Mathematics
Survey Sampling, Vilnius, LSD (2013)
Advanced Methods of Survey Sampling, Neuchatel, Switzerland, Swiss Federal Stat Office (2014)

Experience

Data Scientist (2023-now) Samawatt
Develop and maintain complex trading application for the energy market.

Data Engineer (2022 - 2023)
Write scrapers for a service aggregation site.

Data Scientist (2022) NDA
Used ML (clustering and A/B testing) for marketing (targeted advertisement) in the AWS environment.

Data Scientist Consultant (2020 - 2021) Kyrrex
Consultant for the Trading platform. Trading strategy implementation. Portfolio optimization.

Data Engineer (2020) NDA
Maintain AWS infrastructure using boto3/glue. Wrote news scrapers, and developed a scoring model.

Sr. Data Scientist (2019) NDA
Worked on free dialogue chat-bot and speech recognition problem. Used attention-based seq2seq NN and
DSP. Telegram bot creation.

Data Scientist Consultant (2019) Cryptics
Consultant for the Trading platform. Trading strategy implementation.

Sr. Data Scientist (2018 –2019) AlgoTradeSoft
I was developing an algo trading platform for the cryptocurrency market, back-tester, and analytic
platform.Rewrote mathematical modules from Python to C# for production systems. FX Portfolio
optimization.

Sr. Data Scientist (2018) EXMO crypto exchange
Wrote web-scrapers, and developed a new cashback system for crypto exchange. Provided user analytics such
as clusterization.

Sr. Data Scientist (2017) Trust Sourcing
Two projects: time series forecasting (100k+ different TS) and classification problem in the NLP domain.

Data Scientist (2016 - 2017) Lynx Capital Partners
Worked in high-frequency trading domain. (Q/K language, R, Python)

Data Scientist (2016) TenViz
Startup was oriented on global stock market (daily data). Used R.

Chief Specialist (2012 - 2015) State Statistic Service of Ukraine
Worked in Mathematical Methods, and Metadata Support Unit.
I was responsible for providing various statistical analysis, create complex sample designs, time series
forecasting, and classification. I used all spectrum of modern software for that time (R, Python, SPSS,
SQL, SAS)

Analyst (2009 - 2012) Card Pay Systems
Created VBA scripts for descriptive statistics, analysis, and day-to-day reporting.