Overview

​

Overview

Ocular is built for flexibility and extensibility which gives developers to choose how to implement the logic of their Search platform.

Plugins allow developers to developers to extend Ocular with custom features or integrate with external third party services.

For example if you want to use Open AI or a custom LLM model as an LLM of choice then you can just install the model as a plugin into Ocular.

Ocular plugins run in the same backend process as the core application thus share the same infrastructure, capacity and maintainace. Plugins also have access to services and database within Ocular via a dependency container.

​

Using Existing Ocular Plugins

Ocular has official plugins that cover functionalities such as Large Language Model providers, Document Processors, Vector Embdedders, Vector Databases etc. The current plugins are listed in the next sections.

​

Installing Plugins

To install a plugin in your backend, add the application to the Ocular config file in your backend.

The following config files are currenlty supported

• core-config-local.js - for running Ocular in Docker.
• core-config-dev.js - for running Ocular in development mode.

Set the following in the ```core-config-xxx.js“ plugins section.

plugins: [
   //..
   {
      resolve: `plugin-name`,
      options: {
        // other configurations
        xxx: xxx
      },
    },
]

​

Plugin Configurations

For official Ocular plugins in the Ocular repository, you can find in the README.md a list of configurations. You can also check the official Ocular documentation below.

​

Custom Plugin Development

• Coming Soon