This project is an opinionated framework aimed at simplifying the development of backoffices in TypeScript. By leveraging a modern, serverless architecture and Google Cloud's robust tools, it offers a comprehensive solution for quickly building scalable, efficient backoffice applications.
This framework strongly relies in Google Cloud. Basically because:
- It provides one of the best developer experiences.
- It has unique components like BigQuery and Firestore, which allows great scalability with pay-as-you-go pricing.
The following is a list of the key technologies we plan to use:
- Authentication: Firebase Authentication
- Database: Firestore (Real-time, scalable, serverless). Also in the plan is to suport relational databases like Cloud SQL as we understand that not all applications can be built with NoSQL databases.
- Analytics: BigQuery
- Serverless Functions: Cloud Run
- Scheduling: Cloud Scheduler
- Secrets Management: Secret Manager
- Server Environment: Focused on Node.js to ensure compatibility with Google SDKs, despite an interest in exploring modern solutions like Deno or Bun when they become more mature.
Our framework is designed to include a comprehensive set of features necessary for any backoffice application, which are currently under development:
- Monorepo: Use a monorepo structure to manage the entire project. This allows code sharing at the same time that is imposes strict boundaries between different parts of the application.
- RPC: Implement a robust RPC mechanism for serverless functions, strongly typed and easy to use.
- Secure Frontend: Ensure the frontend is private and not hosted on public CDNs.
- Data Validation: Implement robust validation using tools such as Zod, Yup, or Joi.
- Domain Modeling: Define entities and schemas clearly.
- Intuitive UI: Design an easy-to-navigate menu and user interface.
- Search: Implement powerful search capabilities to easily navigate data.
- Task Scheduling: Incorporate scheduling capabilities for recurring tasks.
- Data Flow Management: Establish clear data flow and dependencies.
- Real-Time Notifications: Provide immediate feedback and notifications.
- Efficient Data Loading: Optimize data loading mechanisms.
- Monitoring and Analytics: Integrate monitoring and analytics for insight into operations.
- AI/Assistants: Leverage AI to enhance user interaction and data management.
- VSCode Extension: Develop an extension for VSCode to streamline the development process further.
- TypeScript: The entire project is written in TypeScript to ensure type safety and a better developer experience. It is the only language that allows us to share types and logic between the frontend and the backend.
- TypeScript tools: Tsup, turborepo, esbuild.
- Frontend: React
- HTTP Server: Thin custom server on top of Node HTTP server. We don't like the global aspect of express where every handler is like defining a global variable, and we like low-level control over the server.
- RPC: tRPC
- Database: Firestore