AI Watch Identifier App

Замовник: AI | Опубліковано: 22.10.2025
Бюджет: 1500 $

I need a polished, production-ready mobile application that recognises wristwatches from photos and returns brand, model, classification, and indicative price. The app must run natively in Java/Kotlin for Android and Swift for iOS, with a modern, vibrant interface that feels consistent across both platforms. Core workflow 1. Image capture or upload, followed by on-device preprocessing to reduce noise and resize before transmission. 2. Secure HTTPS call to my server’s REST endpoint, sending the cleaned image and receiving JSON with the watch’s identity, classification, and price. 3. Display the result with three actionable buttons beneath each watch image: • Reservation • Negotiation • Cancel When a user taps any of these, the username, role (wholesale employee, individual employee, or data entry clerk), date-time stamp, and action taken must be logged for a daily activity report. User & data features • Username required at sign-up, linked to one of the three authority levels. • Search history stored per account and rendered in a simple list. • Side-by-side comparison screen allowing any two previously identified watches to be reviewed together. • Temporary reservation queue that expires automatically if no further action is taken. Reporting A background job should generate a daily CSV or JSON report summarising every “Reservation”, “Negotiation”, and “Cancel” action with full user metadata and timestamps. This file is emailed to an admin address and also remains downloadable from an admin-only screen inside the app. Deliverables • Full Android Studio and Xcode projects with well-commented code. • API interface module (Retrofit / URLSession) ready to point at my existing server. • Basic server stub or Postman collection for testing endpoints. • UI assets and any design files. • One-round handover session to ensure successful build and store submission. Please let me know your estimated timeline, any relevant native apps you have shipped, and how you plan to tackle the image recognition preprocessing.