I need an experienced developer or team to build an AI-powered automated system that can: ⸻ 1. Use AI / Computer Vision to detect empty land plots The system should use AI / machine learning / computer vision to analyze satellite imagery (Google Maps, Google Earth, or other sources) and automatically identify: • empty land plots • undeveloped parcels in urban areas • gaps between houses • areas with no built structures The AI should generate for each detected plot: • GPS coordinates • polygon boundaries • estimated area • classification confidence (If needed, OpenStreetMap building footprints can be combined for better accuracy.) ⸻ 2. Retrieve cadastral data using Catastro Spain API Once a plot is detected by the AI, the system must query the official Catastro API (WMS/WFS/REST) and extract: • Cadastral Reference (REF Catastral) • official area (m²) • land classification • polygon boundaries • coordinates • any additional publicly available cadastral information ⸻ 3. AI-assisted visualization dashboard The system must include a web-based dashboard with: • interactive map (Leaflet / Mapbox / Google Maps) • AI-detected plots displayed as polygons • click to view cadastral information • filters by size, class, confidence level • option to run AI analysis for a selected town or drawn area ⸻ 4. Report generation The system must produce: • CSV/Excel file with full parcel data • GeoJSON export • optional PDF report with: • map snapshots • detected parcels • basic statistics • AI confidence scores ⸻ 5. Technical requirements • Python-based backend • AI/ML or Computer Vision (TensorFlow, PyTorch, YOLO, Detectron, etc.) • GIS libraries (geopandas, shapely, rasterio) • Google Maps / OSM integration • Catastro API integration