1. Building decision support systems based on machine learning and advanced analytics.
2. Financial forecasting using classic econometrics tools.
3. Research and development studies of innovative decision support systems based on deep learning, reinforcement learning, and natural language processing.

Core technology stack:

• Python scientific environment: scipy, NumPy, scikit-learn, pandas;
• Neural networks toolset: Tensorflow, Tensorboard;
• RL and scalable AI tools: Ray, RLLib.

Main tasks:

1. Building predictive models to support business processes in various areas - supply chain management, cost/price analysis, etc. Using multiple machine learning and statistical techniques:

• Time series analysis
• Regression, ANOVA
• Machine learning models (deep learning, gradient boosting, model ensembles)

2. Building optimization engines for supply chain management

3. Preparing research papers and academia workshops related to the Teams’ activity

4. Core technologies:

• Python scientific environment: scipy, NumPy, scikit-learn, pandas
• Optimization engines: cvxpy with different solvers, Matlab
• Neural networks toolset: Tensorflow, Tensofboard, GPU computing
• Databases: MSSQL
• Interactive dashboards: MS PowerBI, Python Streamlit

I run my own small consulting company—complex advisory and consultations on possible use cases for machine learning algorithms in various domains. Additionally:

• building scientific reports and applications on demand
• preparation and analysis of statistical experiments
• conducting training and workshops for teams and firms.

I’m cooperating with commercial companies as well as universities or research labs.

As a Senior Data Scientist, I’m responsible for delivering high-quality analysis and machine learning models for external customers and being an internal consultant and technical leader for some projects.

My main areas of expertise in a Company are applications of:

• neural network models,
• reinforcement learning,
• time series forecasting,
• anomaly detection.

The technology stack includes Python data science toolset (pandas, NumPy, scipy, scikit-learn), neural networks technologies (Tensorflow, Keras, Pytorch), visualization libraries (Streamlit, Dash), and R/Rstudio for statistical/scientific analysis.

Data Scientist in Commercial Management and Business Digitalization Department. Main tasks:

1. Building predictive models to support business processes in various areas - supply chain management, cost/price analysis, etc. Using multiple machine learning and statistical techniques:

• Time series analysis
• Regression, ANOVA
• Machine learning models (deep learning, gradient boosting, model ensembles)

2. Building optimization engines for supply chain management

3. Preparing research papers and academia workshops related to the Teams’ activity

4. Core technologies:

• Python scientific environment: scipy, NumPy, scikit-learn, pandas
• Optimization engines: cvxpy with different solvers, Matlab
• Neural networks toolset: Tensorflow, Tensofboard, GPU computing
• Databases: MSSQL
• Interactive dashboards: MS PowerBI, Python Streamlit

Main tasks:

1. Building predictive models to support business processes in various areas - supply chain management, cost/price analysis, etc. Using multiple machine learning and statistical techniques:

• Time series analysis
• Regression, ANOVA
• Machine learning models (deep learning, gradient boosting, model ensembles)

2. Building optimization engines for supply chain management

3. Preparing research papers and academia workshops related to the Teams’ activity

4. Core technologies:

• Python scientific environment: scipy, NumPy, scikit-learn, pandas
• Optimization engines: cvxpy with different solvers, Matlab
• Neural networks toolset: Tensorflow, Tensofboard, GPU computing
• Databases: MSSQL
• Interactive dashboards: MS PowerBI, Python Streamlit

As the Assistant Vice President of FX Sales Analytics department, developed custom sales analytics application in big data environment for recognizing trading patterns and grouping clients using their portfolio or preferences. Time-series decomposition tools for trend forecasting as well as natural language processing solution – recognizing topics in media feeds.

Technologies:

• Cloudera stack(Hadoop distribution)
• Scala + Spark + Spark Streaming + mllib
• R + caret + mlr + xgboost + H2O
• H2O ai
• Python + scipy + pandas + sklearn
• Pig
• Jupyter notebooks stack

Databases:

• Impala
• Hive
• MS SQL Server
• Informatika Power Center

Delivering analysis of users’ characteristics as well as their post-release behavior. Complex A/B testing and statistical survival analysis (churn/retention). Monte Carlo simulations for predicting future behaviors (reaction to new features). Direct cooperation with Company Management.

Technologies:

• Python + scipy + pandas + sklearn
• R + caret + mlr + xgboost + H2O
• H2O ai
• Jupyter notebooks stack