Igor Stankevich
Verified Expert in Engineering
Deep Learning Developer
Norwalk, CT, United States
Toptal member since March 16, 2020
Igor is a results-driven quantitative development professional with progressive industry experience in commodities, deep learning, computer vision, and data science. He has demonstrated hands-on financial and AI software development experience with in-depth knowledge of Python. Igor consistently attempts to find the best value-add opportunities and contribute to the team as a whole.
Portfolio
Experience
- Python - 12 years
- MongoDB - 7 years
- Scikit-learn - 7 years
- Deep Learning - 7 years
- Amazon Web Services (AWS) - 6 years
- Computer Vision - 5 years
- Quantitative Research - 5 years
- Commodities - 4 years
Availability
Preferred Environment
Jupyter Notebook, Anaconda, Python, Artificial Intelligence (AI), Machine Learning, TensorFlow
The most amazing...
...thing I've developed for a major hedge fund was a business-wide Python modeling framework for quantitative researchers, with cloud risk profiling capabilities.
Work Experience
Head of Pricing and Modeling
Arbol
- Developed risk quantitative framework for derivatives pricing.
- Created and deployed several (20+) derivatives pricing models.
- Led a team of seven quantitative developers, four quantitative researchers, AI engineers, and five pricing analysts.
Senior Data Scientist
CVS Health - Data Strategy, Analytics and Enterprise Analytics
- Developed deep learning models to capture interactions with CVS customers.
- Built a Python framework that unified uplift modeling.
- Developed and deployed uplift models that brought about $10 million of extra revenue.
Senior Quantitative Developer
Citadel
- Designed and developed a modeling and error decomposition framework that enabled interaction across various models and helped to compute errors for each modeling component independently.
- Developed a futures spreads charting tool, that allowed portfolio managers and analysts to quickly select and pivot various combinations of price time series.
- Designed and implemented a backtesting framework core, that let analysts use common strategy performance evaluation metrics.
- Developed web/API scrapes for more than 30 data sources (Pandas, Selenium, beautifulSoup) to feed production models.
- Converted several fundamental models by restructuring existing code, building new components and rewiring them (Pandas, NumPy, Scikit-Learn, statsmodels, lp_solve).
- Organized production model runs scheduling, monitoring, and reporting that helped to bring down forecasts’ availability time.
Biometrics R&D tech Lead
Technoserv
- Developed facial recognition system core with a 99.8% match rate (Chainer/Caffe, Tornado, React), that replaced the legacy system with 75% accuracy and became part of a corporate facial recognition cloud.
- Managed biometrics team of three developers, focusing on Morpho/Safran/L1 solutions integration.
- Oversaw and managed biometrics technology deployment and integration into Russian government agencies.
R&D tech lead
Techno-traffic LLC
- Developed license plates recognition system using proprietary license plates database (Python, OpenCV, dlib, Tornado, AngularJS), that was integrated into federal highway project on 40 toll lanes.
- Managed toll roads software development team of 20 developers.
- Delivered and deployed three layers (embedded, control, and billing) national tolling system.
- Oversaw video-based vehicle classification system development that replaced 3x more expensive embedded systems.
Quantitative Developer
FX AI Investments
- Designed and developed a carry-trade system on JPY crosses (served as the main strategy for two years).
- Implemented more than 50 custom technical indicators for MT4 trading platform (Python, MQL4).
- Developed the backtesting framework (MQL4 + Matlab) that allowed traders to verify their strategies.
Experience
Commodities modeling framework
• Developed complex modeling graphs shared caching strategies (MongoDB/Arctic, Redis, AWS S3) that cut down model run times and enabled quicker collaboration and scenarios exploration between analysts
• Provided support for scenarios management (MongoDB) and execution on a cloud (AWS farm + S3)
• Built an extensive test suite, covering more than 300 functional, integration and performance requirements
• Deployed detailed and intuitive documentation, along with use cases description into centralized system
• Worked with head of directional analysts to design a concept and iterate on core aspects of the framework
Facial recognition cloud
Development steps:
1. Detailed research on most recent technologies available
2. Core model coding based on academia papers, and training on publicly available dataset
3. Proof of concept (Flask web server + Angular UI): recognizing people on video
4. Core library, trained on proprietary faces database (50 mln faces), with tests and documentation
5. Cloud integration API and scaling
License plate number recognition
Development workflow:
1. Solution design: detection approach, manual features, template segmentation and data aggregation approaches
2. Manual and semi-manual labeling with proprietary and public tools, on a private data set
3. Proof of concept implementation: Tornado server + Angular UI, used proprietary online video (brand new real time data)
4. Optimized application for low-performance lane controllers
5. Deployment
Video based vehicle classification system
Futures spreads charting application
• Developed vectorized computation library that speeded up spreads and seasonal strips computation tenfold (pandas, numpy)
• Implemented REST backend with load balancing to interact with UI (Flask, gunicorn)
• Led whole application development by iteratively communicating with UI developer and end users, adding/updating requirements, gathering feedback and presenting new functionality
• Deployed “how-to” guides, along with detailed documentation into centralized wiki
Charts pattern recognition library
Development workflow:
1. Timeseries conversion into graphical charts, with manual and semi-manual patterns labeling
2. Testing variety of deep learning approaches
3. Model training and fine-tuning
4. Core library and application development
5. Flask server deployment on AWS EC2 instances
Strategy backtesting application
Application #1: Strategy testing and optimization framework, with signals generation and performance evaluation on the fly. Framework used Monte-Carlo simulations to pick features' combinations based on the trader-defined universe. These combinations were used to generate series of signals, accounting for patterns' potential biases and create performance report.
Application #2: Core library took holdings and returns series and applied pre-configured metrics to them, generating reports about the strategy, including recommendations based on pre-defined strategy evaluation rules. Library was incorporated into a Flask server, which handled UI requests.
Family Finance App
Data acquisition framework
- web scraping or REST API download,
- data cleaning: missing data removal, outliers and inconsistencies detection, manually pre-configured checks
- data checks/alerts: helped to mitigate risk of biased/incorrect forecasts
Framework added a structure to companies data acquisition strategy, which allowed to systematically approach to most of the data feeds business had.
Approach was tested in production on more than 50 data sources, which were feeding production models.
Framework also provided unified data access API for analysts to work with in jupyter notebooks.
Trading and analysis platform integration
Implementation steps:
1. Research on the integration approaches from MT4 side
2. Socket protocol implementation based on Twisted (python)
3. Load testing strategies scheduling and constant health checks
Pricing Framework
Education
MicroMasters in Finance
Massachusetts Institute of Technology - Boston
Master of Arts Degree in Cognitive Science
Russian State University for the Humanities - Moscow, Russia
Master of Science Degree in Economics and Financial Engineering
Moscow Aviation Institute (National Research University) - Moscow, Russia
Certifications
MIT MicroMasters in Finance
Massachusetts Institute of Technology
Certified AWS Solution Architect
AWS
Scalable Machine Learning
Berkeley on edX
Big Data XSeries
Berkeley on edX
Artificial Intelligence
Berkeley on edX
Skills
Libraries/APIs
Pandas, Scikit-learn, TensorFlow, TensorFlow Deep Learning Library (TFLearn), NumPy, OpenCV, Flask-RESTful, Beautiful Soup, Dlib, Google API, XGBoost, React, Spark ML
Tools
PyCharm, MATLAB, StatsModels
Languages
Python, MQL, Python 3, MQL4, C++, JavaScript, TypeScript
Paradigms
REST, Microservices, Quantitative Research, Management
Platforms
MetaTrader, Amazon Web Services (AWS), Amazon EC2, Anaconda, Jupyter Notebook, AWS Lambda
Storage
Amazon S3 (AWS S3), MongoDB, Redis
Frameworks
Chainer, Flask, Caffe, AngularJS, Apache Spark, MXNet, Spark
Other
Computer Vision, Technical Analysis, Deep Learning, Artificial Intelligence (AI), Machine Learning, Research, Forecasting, APIs, API Integration, Software Development, Commodities, Natural Language Processing (NLP), Financial Markets, Quantitative Modeling, Technical Forex Concepts, Forex Analysis, Tornado, Data Scientist, Generative Artificial Intelligence (GenAI), Data Science, Equities, Oil & Gas, Fundamental Analysis, Generative Pre-trained Transformers (GPT), Random Forests, Decision Trees, Neural Networks, SaaS, Risk, Risk Models, Derivatives, Option Pricing, Quantitative Finance, Development, Quantitative Risk Analysis, Large Language Models (LLMs), Finance, Statistics, Probability Theory, Modeling, Mathematics
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring