Propty is a comprehensive property management platform built for the Dutch rental market. It handles everything from tenant screening and lease management to maintenance requests, financial reporting, and regulatory compliance. With over 420,000 lines of production code, Propty serves real property managers across the Netherlands and is available as both a web application and a native mobile app on iOS and Android.

What makes Propty remarkable is not just its feature set, but how it was built and continues to evolve. Propty was developed and is maintained by AI-native teams, a small group of humans providing strategic direction and architectural oversight while AI agents handle the bulk of day-to-day coding, testing, and deployment. This case study walks through the journey from concept to production platform.

The Challenge

The Dutch property management market is one of the most complex in Europe. Tenants enjoy strong legal protections, regulations change frequently, and property managers need tools that handle everything from rent collection and annual tax reporting to inspection scheduling and repair coordination. On top of that, the Netherlands is a multilingual business environment, property managers and tenants may communicate in Dutch, English, German, or French.

Existing solutions on the market were either prohibitively expensive, too rigid to adapt to changing regulations, or simply outdated. Building a platform from scratch that could compete with entrenched players would traditionally require a large development team, years of work, and millions in funding. For a lean startup targeting a specialized market, a fundamentally different approach was needed.

The requirements were demanding: rental agreement lifecycle management, property inspection workflows with photo documentation, a full repair and maintenance system, invoicing and financial management, multi-language support, and mobile-first design so property managers could work on the go. All of this had to comply with strict Dutch regulatory requirements including GDPR and financial regulations.

The AI-Native Approach

Rather than assembling a traditional engineering team, we took an AI-native approach. AI agents handled the majority of coding, testing, and deployment tasks, working under the guidance of a small human team responsible for architecture decisions, code review, and product direction.

The development workflow was built around continuous delivery. Feature branches were created for every change, validated in a staging environment that mirrors production, and only merged after passing automated checks and human review. This tight feedback loop meant new features could go from concept to production in days rather than weeks.

AI-powered code review became a cornerstone of the process. Every pull request was analyzed for code quality, consistency with existing patterns, and potential issues before a human reviewer even looked at it. This layered approach caught problems early and kept the codebase clean as it grew.

Automated testing was treated as non-negotiable from day one. Unit tests, integration tests, and end-to-end tests were written alongside every feature. The CI/CD pipeline ran the full test suite on every commit, ensuring nothing shipped to production without validation at multiple levels.

Key Features

Rental Agreement Lifecycle Management

Propty manages the full lifecycle of rental agreements, from drafting and digital signing through renewals and terminations. Templates are customizable per property type, and the system automatically tracks key dates like renewal deadlines and rent adjustment periods, sending notifications to both property managers and tenants.

Property Inspections

The inspection module allows property managers to conduct move-in, move-out, and periodic inspections directly from their mobile device. Photo documentation is captured on-site and attached to structured inspection reports. Findings can be converted into repair tickets with a single tap.

Repair and Maintenance Workflow

Tenants can submit maintenance requests through a self-service portal, complete with photos and descriptions. Property managers triage, assign, and track repairs through a unified workflow. The system tracks response times, costs, and contractor performance over time.

Invoicing and Financial Management

Propty generates invoices, tracks payments, and produces financial reports for property owners. Integration with banking APIs enables automated transaction matching, so incoming rent payments are reconciled against outstanding invoices without manual effort.

Banking Integration

Automated bank feed connections pull transaction data in real time. The system matches transactions to tenants and invoices using intelligent rules, flagging exceptions for manual review. This eliminates hours of weekly bookkeeping for property managers handling large portfolios.

Email Management

A unified inbox aggregates emails from connected provider accounts, linking messages to the relevant tenant, property, or maintenance ticket. Property managers can respond without leaving the platform, and the full communication history is preserved for compliance purposes.

AI-Powered Assistant

An integrated AI assistant helps property managers draft communications, summarize tenant histories, and answer questions about regulations. The assistant has access to the relevant context for each property and tenant, making it genuinely useful rather than generic.

Tenant and Owner Portals

Tenants access a self-service portal to view their lease details, submit maintenance requests, download documents, and communicate with their property manager. Property owners get a separate portal with financial overviews, occupancy rates, and maintenance summaries for their portfolio.

Push Notifications and Real-Time Updates

Both the web and mobile apps deliver real-time notifications for important events: new maintenance requests, payment receipts, lease expirations, and inspection reminders. Notification preferences are fully customizable per user.

Technical Highlights

Propty is built on a modern full-stack TypeScript architecture. The frontend uses React with responsive design that adapts seamlessly from desktop to tablet to phone. The backend exposes a RESTful API with comprehensive endpoint documentation, making it straightforward for integrations and future extensibility.

Data persistence is handled by PostgreSQL, chosen for its reliability and support for the complex relational queries that property management data demands. A caching layer ensures that frequently accessed data, like property listings and dashboard summaries, loads instantly even for managers with large portfolios.

Background job processing handles time-sensitive tasks like payment reconciliation, notification delivery, and report generation without blocking the user experience. The mobile apps, built with React Native, share business logic with the web application while delivering a native experience on both iOS and Android.

The automated deployment pipeline takes code from merge to production with zero manual steps. Every deployment is versioned and reversible, and the platform maintains 99.9% uptime through redundancy and health monitoring.

Integrations

A property management platform does not exist in isolation. Propty integrates with a range of external services to provide a seamless experience:

  • Payment processing: Stripe and Mollie for flexible payment collection, supporting iDEAL (the dominant Dutch payment method), credit cards, and direct debit
  • Banking APIs: Automated bank feed connections for real-time transaction import and reconciliation
  • Email providers: Integration with major email services for the unified inbox feature
  • Translation services: Automated translation support across Dutch, English, German, and French for multi-language communication
  • Push notification services: Cross-platform notifications for both mobile and web users
  • Document storage: Secure document management for leases, inspection reports, invoices, and correspondence

Results

The numbers speak for themselves. Propty has grown to over 420,000 lines of production code, a codebase comparable to what a team of 10-15 engineers would typically produce over several years. The platform is live and serving real customers in the Netherlands, with mobile apps available on both the App Store and Google Play.

New features ship continuously. The AI-native approach means that customer feedback can be translated into production features in days rather than sprints. Bug fixes typically go from report to deployment within hours. The platform is maintained and improved on an ongoing basis at a fraction of the cost of a traditional engineering team.

The development cost was approximately 85-90% lower than what a traditional team would have required for the same scope, without compromising on quality, test coverage, or security. This cost advantage is not just a one-time saving; it compounds month over month as maintenance and feature development continue at the same favorable economics.

Lessons Learned

Building Propty taught us a great deal about the practical realities of AI-native development at scale. Here are the lessons we carry forward into every new project:

  • AI agents excel at consistent, repeatable work. Writing API endpoints, building UI components, generating tests, handling migrations, these are tasks where AI agents deliver both speed and quality that matches or exceeds human output.
  • Human oversight remains critical for architecture. The big decisions, data model design, integration strategy, security architecture, still benefit enormously from experienced human judgment. AI agents build well within the boundaries you set, but setting those boundaries is a human responsibility.
  • Automated testing is non-negotiable at this scale. With 420,000 lines of code and multiple deployments per day, you cannot rely on manual QA. Invest in testing infrastructure from day one. The cost is trivial compared to the cost of production incidents.
  • Staging environments prevent production surprises. Every change passes through a staging environment that mirrors production before it reaches customers. This single practice prevents more issues than any other quality gate in the pipeline.
  • Domain expertise cannot be replaced by AI alone. Dutch rental law, municipal regulations, financial compliance, these require real domain knowledge. AI agents are excellent researchers and implementors, but someone needs to know what questions to ask.
  • Documentation is an investment, not overhead. Thorough documentation of architecture decisions, integration patterns, and business rules makes AI agents dramatically more effective. The better the context they receive, the better the output they produce.
  • Start strict, loosen gradually. We began with tight controls on agent autonomy and expanded as confidence grew. This is the right approach for any team adopting AI-native development.

Conclusion

Propty demonstrates that AI-native development is not a theoretical concept. It is a practical, proven approach that can produce large-scale, production-quality software serving real customers with real money at stake. The platform handles complex Dutch regulations, processes financial transactions, and delivers a polished user experience across web and mobile.

The approach is not without its challenges. Quality control, domain expertise, and human oversight remain essential ingredients. But the core promise, dramatically faster delivery at dramatically lower cost with comparable or better quality, is being fulfilled today, not in some speculative future.

If you are considering building a software product and want to explore the AI-native approach, we would welcome the conversation.