Dmitriy Piskarev

Boosted approval rates by 25% and reduced risk by 15% through LGBM modeling, achieving a Gini gain of 21%. Developed an in-house dashboard to monitor model performance metrics and built a YOLO-based ID OCR pipeline, cutting KYC costs by 23%. Led scalable MLFlow deployment pipelines for productionized ML models and managed a team of 11 across Data Science, CV, and Anti-Fraud using Agile tools.

Developed a real-time risk assessment pipeline utilizing Databricks, PySpark, Scala, and SQL. Created Metabase dashboards for continuous model and variable monitoring. Engineered robust data pipelines for large-scale transaction modeling and feature extraction, while also leading internal workshops and guiding company-wide ML initiatives.

Built corporate credit risk models that enhanced Gini performance by 11%. Automated IFRS migration matrix calibration processes. Streamlined variable engineering pipelines to improve predictive accuracy.

Structured multiple M&A deals, built comprehensive DCF models, and conducted thorough asset evaluations. Managed various financial analysis projects.

Created over 100 statistical arbitrage strategies with theoretical annual returns exceeding 20%. Designed low-correlation alpha strategies using financial fundamentals and tensor operations.

Designed advanced robotic systems, including PCBs using Altium, and mechanical assemblies using SolidWorks. Contributed to the development of complex engineering projects.

Completed both Master's and Bachelor's degrees in Robotics and Mechatronics. The curriculum covered various aspects of robotic systems and mechatronic engineering.

Pursued a PhD focusing on Financial Markets and FinTech. The program involved in-depth research and analysis within these specialized areas.