Leading the development of an MLOps platform for Industrial AI in the metals and mining industry. The platform has 7 on-premises installations in Tier-1 ore refining plants with 100+ running ML models and controllers, producing over $130M EBITDA yearly.

Led the development and testing of a defects detection system for FMCG using Computer Vision models, integrating connected devices for enhanced monitoring and control. The main features include real-time defect detection with a size of 0.1 mm on a conveyor with a size of 400 mm. The models were trained on synthetic data processed by generative neural networks to reduce the training data collection time by 90%.

Executed the full development cycle of machine learning model development, including design, testing, and deployment across diverse successful projects in Aerospace, Oil and Gas, Mining, and FMCG industries, with several pilots successfully transitioning into fully deployed products.Developed a Python Framework for the standardization of solutions for industrial predictive maintenance. This development resulted in a 30% reduction in maintenance costs.

Developed software for seismic data analysis automatization, leading to a 50% reduction in time spent on preparation.

Designed and developed a desktop application for the interpretation and visualization of marine magnetotelluric-sounding data. This application streamlined data analysis, improving the efficiency of interpretation processes by 45%.