We created a local AI agent that watches live video feeds, archived video files or still images from video cameras and helps create a record of cars, trucks, buses, people and animals that are moving past a retail store. Designed for the real-estate industry to help justify higher price per square foot rent, the system leverages the YOLO (You Only Look Once) model for a fast and efficient system. We also overlaid weather, holiday events and other demographic information to create insights and analysis (i.e. most truck traffic is between 3-5 pm; car traffic on rainy days is down 35%), and use a local agent to comb through and analyze the data.