Web App for Real Estate

Challenge

We had to develop a web application for operations with real estate without intermediaries. The app is intended to rent, sell, and buy homes, apartments, and other kinds of real estate in Spain, Panama, and Costa Rica.

Solution

The developed web application has a microservice architecture. All microservices are deployed in a Docker container. When the load increases, it’s possible to easily switch to Kubernetes by launching several instances of loaded microservices to improve performance.

The application uses the following microservices:

API-gateway: entry point to the backend of the application; to reduce the number of requests and limit the return of data to the end-user, the GraphQL query language is used.

User-microservice: designed to work with users, appointments, integration with Google Calendar.

Product-microservice: designed to work with ads (CRUD – create, read, update, delete); ElasticSearch is used to search the application quickly.

Billing-microservice: integration of the web application with the Stripe electronic payment processing system.

Cloud-microservice: used for saving images and ads to AWS S3 (or Amazon S3) cloud object storage and adding watermarks.

Notification-microservice: serves for users notification – mail service and Twilio system for SMS notifications.

Common-be: contains functionality that is used across all backend microservices.

For the front end of the application, the Next.js framework was used.

To ensure data security, our team set up the system to check the user’s role before giving access to the information. The database and all configuration parameters of the application are stored on the server where the web application is deployed, making it impossible to get external direct access to information.

Tech Stack: TypeScript, React, Redux, Next.js, Node.js, Nest.js, GraphQL, PostgreSQL