Working as an artificial intelligence developer on a breastcancer research team. The best AI-solution in mammography by medical Moscow leaderboard. I have created several new releases in my team. This allowed us to improve the metrics of our solution, gain a foothold in various ratings, and continue to sell successfully

I was engaged in the creation and maintenance of computer vision solutions related to document recognition, object detection and audio recognition. My solutions have been successfully sold to big companies such as Megafon, DNS and Samolet

Developing an AI-Solution for early detection of breast cancer. The best AI-solution in mammography by medical Moscow leaderboard Detection: FasterRCNN with additional heads Segmentation: MaskRCNN • Create a light-weighted segmentation model for semantic segmentation of breast’s, which greatly accelerated preprocessing (boost several times by speed) • Wrote an OOP pipeline for loading data in database (Postgres) • Processing with a lot of labeled images (Supervisely) • Create a docker-compose pipeline that automated test’s for database • Data collection and generation for segmentation model training • Optimized API working • Rewrite logic of models for working with Postgres(FastApi, sqlmodel, sqlalchemy)

Deploying models, feature engineering, etc