Search Blueprints as Collection Providers (GA)

The ability to use a Search Blueprint as a Collection Provider has moved from Release to General Availability (GA). This integration allows you to leverage the advanced filtering and boosting logic of Blueprints directly within Collections to display highly tailored results.

Key Enhancements

• Opt-in Integration: This remains an optional configuration. Users who previously accessed this via the LPS-129412 Feature Flag will find the functionality now natively available.

• Subtype Filtering: You can now limit queries to specific subtypes directly within the Searchable Types configuration (located under the Source section, formerly Query Settings).

• Simplified Filtering: For Documents and Web Content Articles, you can now select specific Document Types or Web Content Structures within the UI. This eliminates the manual overhead of creating custom query elements to filter by structure keys or IDs.

Semantic Search with OpenAI Integration (GA) [For LES Subscribers only]

Semantic search, powered by direct integration with OpenAI as a text embedding provider, has officially moved from Beta to General Availability (GA).

Configuration & Opt-in

To ensure zero impact on existing installations, this feature remains opt-in. To begin leveraging semantic search, administrators must manually configure the connection:

• Path: Navigate to Instance Settings > Search > Semantic Search.

• Action: Enable the OpenAI Text Embedding Provider and configure semantic ingestion for your desired content types and languages.

• Documentation: Please refer to the official guide for detailed setup instructions and for customizing search with semantic search.

Status of Other Providers

While OpenAI is now GA, other text embedding providers (such as txtai and Hugging Face) remain in Beta. These can still be accessed and tested by enabling the existing feature flag (LPS-122920) under Instance Settings > Feature Flags > Beta.

Assets that are currently being modified within a publication are now protected from deletion in production by default.

Previously, this behavior was optional and disabled by default. It is now enforced automatically to ensure greater stability and data integrity when working with publications.

Key Benefits:

• Prevents most of the deletion modification conflicts that could only be resolved by discarding changes

• Provides safer default behavior without requiring configuration