I want to launch an Android-only application that showcases property listings and layers in practical AI assistance to make the experience smarter for both managers and renters. At its core the app must let a property manager add, edit, and remove listings (images, price, availability, basic details). The AI component should work behind the scenes—think auto-tagging amenities from photos, suggesting competitive pricing ranges, or surfacing the most relevant homes to a tenant based on their previous in-app behaviour. I’m open to the exact technique you use (TensorFlow Lite, ML Kit, embedded models, or a lightweight cloud call), so long as it stays responsive on typical mid-range Android devices. No third-party property-management or CRM integrations are required; the app should run as a self-contained product with its own lightweight backend or local data layer. A clean Material-style UI, Kotlin codebase, and modern Android architecture (Jetpack components, MVVM or similar) will make maintenance easier down the road. Deliverables • Full Android source code and build instructions • Compiled, installable APK targeting Android 11+ • README that documents model choice, training data source (if custom), and how to retrain/replace it • Short video or screenshots that prove the AI tagging or recommendation logic is functioning Acceptance criteria 1. I can create at least one property listing, attach three photos, and have the app auto-suggest tags (e.g., “balcony”, “garage”) from those images. 2. A new user who browses gets property suggestions ranked by relevance after they have viewed or favourited at least two listings. 3. All core flows run offline once data is cached, with any cloud calls failing gracefully. Please outline your proposed tech stack, model approach, and rough timeline in your bid so I can compare options quickly.