AI/ML Engineer, Data Scientist

На данный момент Data Scientist ML Engineer.

Мои компетенции и опыт

AI/ML Engineer and Data Scientist with 5+ years of commercial ML experience and 4 years in mathematical modeling. Built and deployed end-to-end AI/LLM applications and production ML services, covering the full lifecycle from R&D and model evaluation to API development, testing, deployment, and production support.

Key achievement: led R&D and developed the mathematical core of an in-house predictive diagnostics platform that replaced a commercial solution, saving over $5M in licensing costs.

Data Scientist | ML Engineer (Dec нужен доступ к резюме Sep 2026)

SIBUR Digital LLC, Moscow, Russia

• Led R&D and built the mathematical core of a predictive diagnostics platform supporting 500+ ML models across several hundred industrial assets. The platform replaced GE SmartSignal, saving over $5M in licensing costs.
• Built production gRPC services for predictive diagnostics and forecasting; optimized ML algorithms and migrated REST to gRPC, reducing response time from up to ~1 s to ~50 ms. Developed Protobuf APIs, automated tests, PostgreSQL integration, Docker deployment, and GitLab CI/CD.
• Built and deployed a production LLM-powered diagnostic assistant for incident analysis, time-series interpretation, report generation, and maintenance recommendations, reducing estimated report preparation time from ~30 to ~5 minutes.
• Mentored two interns whose contributions became part of the core product, providing technical guidance and code reviews.
• Built a RAG pipeline for enterprise knowledge retrieval from Confluence, achieving a 90% correct-answer rate on an expert-evaluated test set.
• Developed an NLP approach for early warning of Vertica database failures, validated on two real-world cases; results led to two peer-reviewed research papers.
• Improved F1-score by 20 percentage points in company-name fuzzy matching for lead generation through similarity feature engineering and gradient boosting.

Tech stack: Python, PyTorch, scikit-learn, LightGBM/CatBoost, LangChain, gRPC/Protobuf, REST API, SQL, PostgreSQL, FAISS, SHAP, Streamlit, n8n, Docker, GitLab CI/CD.

Senior Specialist (Sep нужен доступ к резюме Dec 2021)

IGiRGI JSC (Rosneft), Moscow, Russia

• Contributed to the development of a patented technology and software solution for predicting reservoir rock fracturing, aimed at reducing financial losses during drilling.
• Built a Python application using Dynamic Time Warping (DTW) to identify similar segments in gamma-ray logs and detect signs of the well trajectory exiting the oil-bearing reservoir interval.
• Developed mathematical models and computational algorithms for oil production using Python and Fortran, including processing and analysis of large geological datasets.

Achievements: co-author of a patented technology; 2nd place in Rosneft's IT Marathon hackathon; publications in industry research journals.

Tech stack: Python, Fortran, MATLAB, SQL, DTW, numerical methods.