Managing Semantic Search Embedding Models
Embedding models convert indexed text and user query text into vector embedding. PeopleSoft uses these embedding to support semantic matching in OpenSearch.
Use the Embedding Models page to view the models that are imported or available for Semantic Search. This page identifies the model that PeopleSoft uses to generate embedding for indexes enabled for semantic search.
The page can include delivered models and custom models that are registered and deployed in OpenSearch and then imported into PeopleSoft.
Access the Embedding Models page by selecting .
This example illustrates the fields and controls on the Embedding Models page. You can find definitions for the fields and controls later on this page.

| Field or Control | Description |
|---|---|
|
Model Name |
Displays the name of the embedding model imported for Semantic Search. The model name identifies the model used to generate vector embedding for indexed text and search queries. PeopleSoft delivers models that are preconfigured during OpenSearch installation. |
|
Model ID |
Displays the unique model identifier assigned to the embedding model. PeopleSoft uses this identifier to reference the model during semantic indexing and search processing. |
|
Dimension |
Displays the vector dimension generated by the embedding model. The dimension must match the vector mapping used by the semantic-search index. |
|
Active Model |
Select this option to activate the model for Semantic Search. The active model generates embedding for indexes that have Semantic Search enabled. Depending on the PeopleSoft environment configuration, you can switch the active model as needed. Note that changing the active model requires the data to be re-indexed. |
|
Deploy |
Select to deploy the embedding model so that it is available for generating vector embedding. |
|
Save |
Select to save changes made on the page, such as the selected active model. |
|
Import |
Select to import available embedding model information into PeopleSoft. |
You can configure and import custom models using this page. For more information, see Installing Custom Embedding Models.