Skip to main content

Architecture overview

Metadata Capture is an open-source platform that you can deploy in your organisation's infrastructure. This self-hosted deployment model gives you complete control over the application. Before you begin deployment, understand the application architecture so you can prepare your environment.

The Metadata Capture application is a modular platform that integrates several components to enable secure, scalable dataset management. The following figure illustrates the high-level architecture of the system.


Metadata Capture integrates with your organisational infrastructure and is designed for flexible deployment. Here’s what you need to know before you begin deployment:

  • Core application: The Metadata Capture application includes both a frontend (user interface) and backend services, designed to run on your infrastructure. Your users interact with the system through their web browser, which communicates directly with the application.

  • Authentication & access control: The application relies on an external OpenID Connect (OIDC) provider to manage authentication and user permissions. Your organisation must have its own Identity and Access Management (IAM) system, which controls access and role assignments within Metadata Capture.

  • Centralised database: Metadata Capture stores and retrieves all application data using a PostgreSQL database. Your organisation must provide its own PostgreSQL instance and grant Metadata Capture secure access to the database.

  • Email notifications: Metadata Capture includes a notification feature that sends emails to relevant users during key events of the dataset lifecycle. For example, it sends an email when a dataset is submitted for validation, or when it is approved for publication. Your organisation must configure its own SMTP server to deliver these notifications.

Now that you understand the architecture, you can prepare your environment and deploy the Metadata Capture application.