> Developing algo-trading strategies for financial markets (Rust).> Architected and implemented large-scale pipeline for walk-forward portfolio optimisation extensively used by analytics department (Python, Dagster, Polars).> Training ML models to predict trading metrics of strategies' configs (CatBoost).

> Published two A* and Q1 scientific papers in the field of CV and NLP.> Team-lead of NLP research team under supervision of Ivan Titov (top 2% highly-cited NLP scientist in the world).> Developed deep learning framework "XCBs" for cross-modal (images and text) interpretability through conceptualisation (approach was presented in EMNLP-23).> Developed service generating textual structured reports for chest X-Ray images (PyTorch, transformers, nltk).> Developed scalable deep learning framework "spiner" for spine MRI pathology examination (PyTorch).> Implemented deep learning package "purelung" for chest X-ray pre-processing in industrial data pipeline (PyTorch, OpenCV, ONNX).