If you're new to Vertex AI Search, consider following the Get started with media recommendations quickstart tutorial to create a sample app.
Set up a Google Cloud project, turn on Vertex AI Agent Builder, and set up access control for your project. You can use an existing Google Cloud project if you have one already.
Actions
- Complete the steps in Before you begin.
Prepare your data for importing to Vertex AI Search.
Media recommendations apps require two types of data:
Structured media data. Upload metadata information about your media content, such as titles, descriptions, and URIs to the location of your media. Vertex AI Search provides a predefined schema for media. Alternatively, you can use your own schema.
User events. Recording user events is required for media recommendations. User events are required to train your app and to generate recommendations.
Actions
Review information about media data and data stores and prepare your data according to the required schemas and fields in About media documents and data stores. If you are using your own schema, also see Example schema as a JSON object and structured data.
Review media user event requirements in About user events.
Create an app and data store and then import your media data and user events.
How you import media data depends on where you're importing it from. For example, if your data is in Cloud Storage, you can import it using the console or the API by providing the bucket location of your data.
Actions
Bulk import historical user events so that your app can start training.
You can update field settings to make them filterable and filter your recommendations results using those fields.
Actions
To let Vertex AI Search use a certain field for filtering recommendations, set that field as filterable. See Configure field settings.
Apply serving configurations to demote and diversify media recommendations. See Create and manage serving configs.
You can preview your recommendations to check if your recommendations are appearing as expected.
Actions
To preview your recommendations, use the Agent Builder console or the API.
Console. Use the console Preview page to preview your recommendations. See the Console instructions for the kind of data that your app uses in Get recommendations.
API. If you're integrating API calls into your application, make API calls to preview your recommendations. See the REST instructions for the kind of data that your app uses in Get recommendations.
When you're happy with the preview version of your media recommendations app, share it with your users by deploying it to your website.
Actions
To deploy your recommendations app, integrate API calls into your server or applications. For more information about making API calls, see the REST instructions for the kind of data that your app uses in Get recommendations.
For client library resources, see Vertex AI Agent Builder client libraries.
You can maintain your app to ensure that latest and necessary data is available in your data store.
Media recommendations apps are automatically tuned every 3 months. However, if your data store undergoes significant change or if you do a bulk import of user events, you should manually retrain your app.