Description
** Please make sure you read the contribution guide and file the issues in the right place. **
Contribution guide.
Is your feature request related to a problem? Please describe.
Currently, the Vertex AI Search API in the ADK accepts datastore, engine, maxResults, and filter as parameters. However, it lacks the ability to specify dataStoreSpecs[] and boost_spec. This omission is problematic for users who need fine-grained control over their search queries, especially when working with Engines that have multiple data stores or when they want to influence search results through boosting.
Specifically, without dataStoreSpecs[], it's difficult to configure how specific data stores within an Engine should be searched. This limits the flexibility and precision of search queries in multi-datastore environments.
Furthermore, the absence of boost_spec prevents users from applying boosting rules to their search queries. This means they cannot prioritize certain documents or content based on specific criteria, which is a common requirement for enhancing search relevance and user experience.
Describe the solution you'd like
I would like the Vertex AI Search functionality within the ADK to be updated to include support for dataStoreSpecs[] and boost_spec as parameters.
The dataStoreSpecs[] parameter should align with the specification provided in the Cloud Generative AI App Builder documentation: https://cloud.google.com/generative-ai-app-builder/docs/reference/rpc/google.cloud.discoveryengine.v1#datastorespec. This would enable users to define specific configurations for individual data stores when using Engines with multiple data stores.
The boost_spec parameter should allow users to define boosting rules as described in the Vertex AI Search documentation. This would empower users to promote or demote certain documents based on their content, metadata, or other relevant attributes, thereby improving the relevance of search results.
Describe alternatives you've considered
I tried tampering requests in before model callback but it doesn't allow me to add datastorespecs.
Additional context
The inclusion of dataStoreSpecs[] and boost_spec would bring the ADK's Vertex AI Search capabilities closer to the full functionality offered by the underlying Vertex AI Search API. This would greatly enhance the power and flexibility available to developers building applications with the ADK, allowing them to create more sophisticated and relevant search experiences. This aligns with the goal of providing a comprehensive and robust toolkit for interacting with Google Cloud services.