Vadim
Akhmetov

Team Lead, Data Science — Digital Finance International, Belgrade

MLOps on ClearML: infrastructure for model deployment to production with monitoring and alerting. Scoring models for microfinance and installment-plan organizations, from proof of concept to production maintenance and model evolution. Lead a team of four data scientists: task decomposition, development planning, mentoring.

Senior Data Scientist — RANEPA (part-time), Moscow

Multi-service ML applications with a FastAPI backend. Handwritten Text Recognition (HTR) system for historical documents of the 17th–19th centuries. Interactive educational web applications on Streamlit.

Senior Data Scientist — Ozon, Moscow

Recommendation algorithms for short videos (OzonMoments). ETL on Spark scheduled with Airflow. Video embeddings from deep learning models powering a real-time interest engine. Session-based recommendations with DSSM, improving engagement metrics.

Senior Data Scientist — Rosbank, Moscow

Credit scoring quality: photo scoring for probability of default, OCR for property-rights documents, default scoring on transactional data, external anti-fraud, Auto-ML. Led a team of three data scientists.

Data Scientist — Sberbank, Moscow

Data markup aggregation, employee attrition models, headcount forecasting, OCR models (segmentation, text detection, text recognition), face defect detection.

Computer Vision Engineer — Algomost Ltd., Moscow

Classical and neural video analysis algorithms for manufacturing process automation.

Teacher / Project Manager — Educational Center «Sirius», Sochi

• Safety compliance monitoring: helmet detection with pose estimation and OpenCV color filtering.
• Python and IoT: students built a web service (Flask, SQLAlchemy, Arduino) monitoring pipe sensor signals.
• Computer vision: YOLOv3-based cucumber detection guiding a CNC harvester.