Deliverables
- End-to-end UI/UX Design System
- QA Audit & Technical Assessment
- UI/UX Corrections & Enhancements
- Cross-Platform Mobile Apps (iOS & Android)
- RESTful Backend (Node.js / TypeScript)
- AI Image Processing Pipeline (YOLO + GPT Vision)
- AI Chat Assistant (“Dena”) using LangChain
- Subscription & Monetization System (RevenueCat)
- Multi-Currency & Multi-Region Support
- Offline Capture & Sync System
- PDF & Excel Report Generation
- Push Notification System (Firebase FCM)
- Post-Launch Support & Maintenance
Team Composition
- Duration: May 2025 – Ongoing
- Total Effort: ~395 Engineering Hours
- Team Size: 4 Specialists
Team Composition
- UI/UX Designer
- QA Engineer / Technical Auditor
- Mobile Developer (React Native / Expo)
- Backend Developer (Node.js / TypeScript)
The Challenge
Product Challenges
Traditional property inspection methods are often:
- Expensive and time-intensive
- Static and document-heavy
- Lacking interactivity and real-time insights
- Limited in cost transparency
- Difficult to compare across multiple properties
Buyers required a modern solution that could provide:
- Instant defect detection
- Accurate, localized repair cost estimates
- Actionable negotiation insights
- Multi-property comparison
- Offline functionality for low-connectivity environments
Delivery Challenges
After initial design delivery, the product built by another team presented several issues:
- Critical functional defects
- Missing core features
- Inconsistent UI/UX implementation
- Incomplete AI integration
- Weak architectural foundations
Ropstam needed to recover and enhance the product without restarting development from scratch.
The Solution
Structured Recovery Strategy
Ropstam adopted a phased recovery approach to restore and deliver the product efficiently:
Phase 1: Audit & Gap Analysis
- Functional validation against design specifications
- Codebase and architecture assessment
- Identification of missing and incomplete features
- Cross-platform consistency checks
Phase 2: Stabilization & Alignment
- UI corrections to match approved designs
- Backend optimization and API fixes
- Architectural restructuring
- Completion of critical features
Phase 3: Feature Completion & Enhancement
- AI detection pipeline integration
- Subscription and billing system
- Offline inspection capability
- Multi-currency and localization support
- AI-based issue grouping
This approach enabled Ropstam to salvage usable components while resolving major deficiencies, avoiding a full rebuild and saving both time and cost.
the Design
Development
Launching
Key Features
AI-Powered Defect Detection
The platform uses a hybrid AI pipeline:
- YOLO-based object detection
- GPT Vision for classification
- Severity and confidence scoring
This enables automated detection of property defects directly from images.
Localized Cost Estimation
Each detected issue includes:
- The estimated repair cost ranges
- Region-specific calibration
- Multi-currency support (USD, CAD, GBP, AUD, NZD)
AI Issue Grouping
The system intelligently consolidates related defects across multiple images into unified insights, reducing duplication and improving report clarity.
Inspection Reports
Two levels of reporting:
- Room-Level Reports: Detailed defect analysis
- Property-Level Reports: Aggregated insights and cost summaries
Available in:
- In-app view
- PDF export
- Excel export
Dena – AI Chat Assistant
Built with LangChain and GPT models, Dena enables users to:
- Ask contextual questions about defects
- Upload images for analysis
- Receive region-specific insights
- Maintain conversational continuity
Offline Inspection Mode
Designed for real-world usability:
- Local data storage
- Automatic sync on reconnection
- Background processing and retry handling
Subscription & Monetization
Implemented via RevenueCat:
- Multiple pricing tiers
- Cross-platform validation
- Server-side subscription enforcement
Notifications & Authentication
- Firebase Cloud Messaging for real-time alerts
- Apple Sign-In, Google OAuth, and email authentication
- Fully compliant with platform policies
Tech Stack
Frontend
- React Native (Expo)
- TypeScript
- Zustand
Backend
- Node.js (v20+)
- Express
- MongoDB
- PostgreSQL (pgvector)
- LangChain
Infrastructure
- AWS S3
- Docker
- Heroku
- OpenAI GPT
- Roboflow YOLO
- Firebase FCM
- RevenueCat
Project Timeline Highlights
- May – June 2025: UI/UX Design
- August 2025: QA Audit of Existing System
- Aug – Sept 2025: Stabilization & Feature Completion
- September 2025: Subscription & Offline System Implementation
- Nov – Dec 2025: Post-Launch Enhancements
- Nov 2025 – Ongoing: Support & Maintenance
Results
Android Application
iOS Application
Ropstam successfully transformed an unstable, partially developed product into a production-ready platform.
Key Outcomes
- Successful iOS and Android launch
- ~395 hours of focused engineering effort
- Full AI pipeline implementation
- Live subscription system
- Offline-first inspection capability
- Multi-country scalability from a unified backend
- Recovery of the inherited system without a full rebuild
Business Impact
GreenTag delivers significant value to modern property buyers:
- Enables independent, AI-assisted inspections
- Provides transparent and localized repair costs
- Strengthens buyer negotiation strategies
- Supports comparison across multiple properties
- Establishes a scalable, subscription-based revenue model
- Positions the platform for global PropTech expansion
Client’s Testimonial
The team communicates very well. It’s very refreshing Ropstam suggested innovated that we haven’t thought about. Overall outcome is highly appreciated.


