Predictive modeling data scientist experience
Probabilistic methods & Bayesian inference
- Bayesian inference (hierarchical models), Bayesian Networks, Gaussian Processes (incl. GP hazards, monotone links)
- MCMC (HMC/NUTS), SVI
- Diagnostics & calibration (R‑hat/ESS, reliability/Brier, isotonic)
- Conformal prediction & uncertainty quantification
PPLs & libraries
- PyTorch. TensorFlow, scikit‑learn, PyMC, TensorFlow Probability, Stan, pycox / scikit‑survival
Time series & forecasting
- Time‑series forecasting for hazards/risk
- State‑space models & filtering (e.g., Kalman/particle)
Data & scale
- Wrangling large messy clinical/claims/OCR’d datasets, Missing data & censoring (MICE/IPCW/etc.)
- Feature stores & evaluators (Feast or custom)
- Distributed compute (Dask/Spark)
Model engineering & reproducibility
- Reproducible pipelines (MLflow/DVC/W&B)
- Testing: data/feature‑leakage controls
Domain knowledge
- Survival modeling in life insurance/life settlements, How survival PDFs feed policy DCF/NAV & fund constraints
- Actuarial basics (mortality tables, DCF/NAV)
- Finances, Retail, Real Estate
Commercial Experience
Prompt Engineering and Data Science Analyst, NDA
2024 - now
DATA SCIENCE & ANALYST, Real Estate
May 2021 - 2024
Forecasting of real estate pricing:
- Regression analysis for trends
- Moving windows to consider mostly the newest history
- Seasonal analysis
- Microcycles and large- and medium economical cycles analysis using Fourier approach and patterns recognition
Technologies: Python, python libraries: pandas, numpy, sklearn, scipy, seaborn, matplotlib, plotly
DATA ANALYST, AMAZON FBA PROJECT
July 2020 - April 2021Â
Overview: Next-generation consumer goods company reimagining how the world’s most-loved products become accessible to everyone. We use a deep understanding of rankings, ratings, and reviews to identify and acquire quality brands and use world-class expertise and data science to make their products better or create new ones to meet changing customer demand.
Responsibilities:
- Development of Analytical Dashboards
- ETL implementation for returns and orders data
- Data cleaning, validation, reconciliation
- Preparing of BI Reports using Periscope based on AWS
- Building causal impact estimation for significant sales dynamics changes
- Technologies: Periscope, SQL, Python (Pandas, NumPy, Plotly), R (CausalImpact)
- Create and support ELT data pipelines built on Snowflake and DBT while ensuring high-quality data
- Develop and deploy data warehousing models, and support existing processes/ETLs (extract/transform/load), and functions (in Python/SQL/DBT) in a cloud data warehouse environment using Snowflake, AWS servicesÂ
- SQL statements and developing in Python, Design and develop data pipelines (DAGs). Automation tests.Â
Technologies:Â Sisense BI, AirFlow, ETL, ElasticSearch, Snowflake, Python, SQL, DBT, Pandas, AWS S3, Medallion Architecture, MySQL, Hadoop, Spark, GitLab CI/CD, Kubernetes, LDAP, Automation Test, Pytest, Snowflake Schema, Dimensional Modeling, ER Diagrams.
DATA SCIENTIST, DEVELOPER & PROJECT MANAGER
March 2018 – 2020
Responsibilities:
- Development of Analytical Platform using Nonlinear Dynamical approach and NLP combined
- Preparing of BI Reports using Periscope based on AWS
- Development, implementation, support and enhancement of forecasting systems for cryptos, metals and currencies basing on nonlinear dynamics models; creation of own nonlinear forecasting algorithms
- Work on Adverts parsing and processing projects
- Development of interactive 3D Globe for Economic Data Visualization
- Developing Web applications on ASP.NET Core – e.g. Web API, EF for the following systems:
- Medical data storage system
- Financial flows system
- Global hotels booking system
- Development of interfaces for booking portals
- Different projects for website optimization and development
Technologies: Python (Pandas, NumPy, Plotly, NLTK, Matplotlib, SciPy, TensorFlow, Dash, Flask, etc.), R (Shiny, dplyr, tidyr, caret, rEDM, etc.), C#, Elixir, JavaScript, Microsoft SQL Server
SENIOR MATHEMATICIAN, DATA SCIENTIST
August 2016 – February 2018
Responsibilities:
- Development, implementation, support and enhancement of forecasting systems for sales indices
- Creation of analytical and forecasting products basing on multi-dimensional nonlinear forecasting, multi-models forecasting systems, nearest neighbors, ridge regression, logistic regression, random forest, cluster analysis, factor analysis, principal component analysis
- Preparing multiple decks on the project results for the clients
Technologies: R (Shiny, caret, rEDM, etc.), Python (Pandas, NumPy, Matplotlib, SciPy), PostgreSQL
PROJECT MANAGER, ECONOMIC PROCESSES MODELING EXPERT
March 2008 – August 2016
Responsibilities:
- More than 30 successful industrial market research projects: market research and modeling, SWOT and COST analysis, market niches identification, market volumes and sales forecasting, etc.
- Development and implementation of forecasting and classification systems for a number of subscribers for information & analytical products
- Development and implementation of a nonlinear forecasting system for price forecasts
- Construction of optimization models for industrial markets for deep analysis of structure and prospects of the markets
- Development and implementation of the unique charts format (including Cost Curve and others)Â
- Participation in international conferences
Education
Oles Honchar Dnipro National University.
- PhD (ABD) in Mathematics (Statistics & Probability Theory), 2009 - 2013.
- Master's degree in Mathematics (Statistics & Probability Theory), 2003 – 2008;
- Economist (Enterprise Finance, Bachelor's degree) 2006 - 2009.
