Artem Matiash

A Python-based platform for simulating trading strategies' historical performance. I redesigned and improved the existing single-asset solution to increase the number of assets, timeframes, and data types utilizing the subscription approach and distributed data sourcing. Authored various internal order types such as limit, market, stop-limit, and one-cancels-all orders and library for iterative calculation of 100+ technical indicators and features. Backtester was also integrated into the R&D cycle by connecting the input data stream to a high-speed database and output data stream to the web-hosted dashboard.

A Python-based modular solution for day trading cryptocurrency derivatives based on the readings from multiple exchanges. I developed connectors between the exchange's data source and high-speed database to compress, store, and process data. Processed data was then passed to a signal generator that utilized the exchange's API for optimal order execution. Logging was performed via a chatbot in Telegram. The bot was traded live with test capital to check a researched hypothesis.

A high-speed, data-intensive algorithmic trading strategy that uses trade data to generate signals calculates trend direction and position size and automatically executes orders using exchanges API. Trend and size quantities are based on a pre-trained Machine Learning algorithm with features selected by feature importance and expert knowledge
Estimated Sharpe ratio of 2.5. Traded live on small-cap for half a year.