Developed multiple automated visual quality inspection systems based either on deep learning and classical computer vision. Systems are optimized for deployment on edge devices which ensures constant up‑time and increases security of data. Clients are satisfied and report reduced downtime in manufacturing. Worked on drone vision systems. Their tasks ranged from detection of warehouse items to using SLAM‑assisted navigation. QR, barcode and ArUco marker readers are used to identify each object.Created custom datasets using tools such as Roboflow. Utilized data preparation, feature engineering, data annotation and data augmentation, keeping best practices in mind. Evaluated machine learning models, expanded datasets and modified model architecture to achieve desired results.

Worked part-time alongside college.Computer vision, AI/ML, Python software developer. Visual quality inspection and drone vision systems.

Developed classical and ML computer vision algorithms for detecting faults on MDF boards. Creating custom datasets and using YOLOv5 for quality inspection in industry. Testing MaskRCNN on synthetic datasets.

Unity game engine. Blender. 3D physics.