● Developed and maintained object detection, segmentation, and ocr models for shelf monitoring, cash drawer recognition, vehicle detection (cars and motorcycles), and table cleanliness classification. ● Leveraged internal training apis to launch, monitor, and manage model training workflows. ● Investigated and resolved client-reported recognition issues, improving model reliability and reducing errors. ● Built and optimized data pipelines using python, pandas, numpy, and custom augmentation workflows. ● Conducted experiments and hyperparameter tuning using kubeflow, aws, katib, mlflow, and tensorboard. ● Collaborated with project managers and commercial teams to translate business requirements into scalable machine learning solutions.