Data Scientist2018 - 2019Quoine
Technologies: Git, Bash, PostgreSQL, SQL, IPython, Bokeh, Keras, Scikit-learn, SciPy, Pandas, Python
- Built an adaptive market-making strategy for liquid crypto markets, which included risk control, fair price estimation and prediction of bid-ask spread related market metrics.
- Adjusted a Kalman filter model for FX rates in multi-market executions.
- Designed and developed a flexible backtesting framework for testing and optimization of high-frequency trading strategies; developed effective visual reports of strategy performance.
- Optimized an ETL pipeline and reporting system for analysis of trading activity on different external exchanges, redesigned daily and monthly auto-generated P&L reports.
- Performed research and prototyping for improving predictions of various market metrics (Linear models, decision tree methods, GARCH, recurrent neural networks).
- Led the project to support internal and external audit requests: ad-hoc analysis, reporting and data investigations, data issues backtracking and coordination of required fixes.
Data Analyst (Contractor)2017 - 2017Soft Retail
Technologies: Git, IPython, PostgreSQL, Scikit-learn, Bokeh, Pandas, Python
- Identified fraud cases in transaction and customer data using unsupervised anomaly detection algorithms (local outlier factor, and isolation forest).
- Designed a database scheme and implemented a regular transfer of client data from local CRM to relational DB (PostgreSQL).
- Developed a set of effective indices and indicators for analysis and visualization of profit trends, customer segments, and customer behavior trends.
- Implemented and deployed an interactive online dashboard to track main performance indices.
Quantitative Analyst2009 - 2015Applied Technologies
Technologies: MATLAB, Statistics, Microsoft Excel, SQL, Java
- Developed a set of metrics for investor competence rankings based on SEC13F filings.
- Applied data mining techniques (visualization, decision trees, and clustering) to identify a group of stocks and main trading patterns useful for the generation of buy/sell signals.
- Developed and backtested a set of index strategies; each strategy aimed to cover a specific decision logic, for example, build a balanced portfolio based on top competent investors and so on.
Lecturer Assistant2007 - 2010Chelyabinsk State University
Technologies: MATLAB, Fortran, Java
- Organized practical and laboratory classes (including econometrics, probability theory and statistics, and Monte-Carlo techniques).
- Contributed to a few research projects devoted to statistics and mathematical modeling.
- Developed study materials and home projects for the course "Applied Probability Theory and Monte-Carlo."