AI / Computer Vision Developer Needed for Satellite Image Analysis Prototype

We are looking for an AI / computer vision developer to help us build a prototype that can analyze satellite or aerial images of a property and identify visible surrounding features. The goal is to pass satellite images into an AI model and have the system help identify items such as: ● Buildings or structures ● Parking lots ● Paved areas ● Open land ● Possible changes in site conditions ● Other visible property or land-use features We would also like the system to compare satellite images from different years and highlight meaningful changes between the two images. This may include things like: ● New buildings or structures ● Removed buildings ● Expanded parking lots ● Changes in pavement or developed areas ● Significant land-use changes around the subject property We are open to different technical approaches, including computer vision models, image segmentation, object detection, geospatial analysis, or using existing AI vision models where appropriate. The first phase would be a small proof of concept using a limited set of sample satellite images. We want to understand what is technically possible, what accuracy can reasonably be expected, and what the best workflow would be before building a larger system. Ideal experience: ● Computer vision / image recognition ● Satellite, aerial, or geospatial imagery ● Object detection and/or image segmentation ● Change detection between images ● Python experience ● Experience with APIs or AI model integration ● Ability to explain tradeoffs clearly Deliverables for Phase 1: ● Review sample satellite images ● Recommend the best technical approach ● Build a working prototype that can analyze one image ● Build a comparison workflow for two images from different years ● Provide a short explanation of limitations, accuracy expectations, and next steps Please include examples of any similar work you have done with satellite imagery, aerial imagery, computer vision, object detection, or change detection.