Grigory Evko Email & Phone Number

Dubai, United Arab Emirates

• Developed a state-of-the-art model for time-series prediction, achieving 64% accuracy for short-term price forecasts, vastly outperforming existing benchmarks.
• Built a quantitative trading platform using tick-level data, integrating reinforcement learning for algorithmic trade optimization, resulting in 20% improvement of trade execution efficiency
• Orchestrated a distributed RL infrastructure on a cluster of 64 GPUs, improving model training scalability, and reduced training time by 94% through strategic AWS EC2 cloud optimizations.
• Designed Uniswap v3 liquidity strategies with projected 40-50% APY gains, created a cross-exchange Python simulation environment, and engineered Solidity smart contracts for strategy execution.
• Engineered robust and efficient Solidity smart contracts, paired with a Python backtesting suite, to refine strategy execution and data workflows.
• Conducted in-depth protocol analysis of Uniswap v3, identifying critical shortcomings and conceptualizing an innovative decentralized exchange.

Moscow, Moscow City, Russia

• Designed a Nextflow and Python-based data pipeline for DNA-RNA sequence analysis, enhancing efficiency by 530% andsignificantly reducing computational costs.
• Administered a large-scale genetic database (14TB), implementing data management practices that ensured high data qualityand streamlined access for analysis purposes.
• Applied advanced statistical modeling and machine learning techniques to multi-species genomic datasets, providing keyinsights that supported scientific research and publication.
• Led interdisciplinary workshops in bioinformatics, promoting the adoption of data science methodologies and tools acrossresearch teams.

Moscow, Russia

• Enhanced customer support chatbot accuracy by 13% through advanced NLP model development and iterative training.
• Reduced model training time by 70% by experimenting with and implementing efficient NLP architectures.
• Automated data markup ETL processes, cutting data preparation time by 40%
• Deployed ML models into production, achieving 99.94% uptime and improving service reliability.
• Streamlined Chatbot Classification Pipeline, resulting in a 9.5% increase in classification precision.
• Executed codebase refactoring and MLOps integration to bolster system scalability.