Developer (AI Training/Model Development): ParkingSpotter Project
Responsible for training a custom YOLOv8 model using real-time street camera images from NYC DOT feeds. Built and maintained an automated data pipeline to collect, preprocess, and annotate images for parking status recognition on the ParkingSpotter app. Oversaw the evaluation and refinement of model accuracy through iterative annotation and testing cycles. • Custom image annotation workflow developed for street parking detection • Leveraged YOLOv8 object detection for practical user notifications • Integrated external data APIs and managed data ingestion pipeline • Conducted ongoing evaluation and labeling quality assurance.