At ITMO University, I developed my MSc thesis on real-time 3D perception for autonomous ground vehicles. My training-free pipeline combined YOLOv8m-seg, calibrated LiDAR projection, clustering, and 3D bounding-box fitting; on KITTI, it reached 50.51% Car IoU and ran at about 11 FPS.
I also built a multi-agent study and coding assistant with a hallucination-detecting reviewer. In evaluations on SWE-bench Lite, the best tested cell improved the judge score from 0.511 to 0.710.
My projects include evaluating citation reliability in small open-weight language models and implementing NeRF pipelines for 3D scene reconstruction. I’ve also worked on robot navigation with A*, DWA, and MPC, and on forecasting for robot-assisted environmental monitoring.
Alongside my studies, I served as a Lab Engineer (Lab Demonstrator) at ITMO University, supporting and guiding undergraduate engineering coursework. Earlier, I assessed wind turbine data for predictive maintenance at Sapphire Renewables (Wind Power Company Ltd.).
