Creating recommendation Apps on Agent Builder via Terraform (SOLUTION_TYPE_RECOMMENDATION)

Hi! I was investigating how to create different app types in Terraform for Google Cloud’s Agent Builder and I’ve hit a roadblock regarding changing app types in IaC. All apps are created with the “Search” type, but I need an app of type “Recommendations”.

Use Cases

Creation of Agent Builder Apps that deliver recommendations to users, in addition to search.

350501374-26ce9aed-19e6-42da-baee-12f9a2bb20e1

Attempted Solutions

There are no parameters that allow me to modify the engine’s solution type directly. I’ve tried importing a recommendations app into my Terraform state, but every apply a replacement gets triggered, modifying INDUSTRY_VERTICAL and search parameters, but not app type explicitly.

Proposal

Have a solution_type field with value choices allowed by the API. (https://cloud.google.com/generative-ai-app-builder/docs/reference/rest/v1/projects.locations.collections.engines.

Solution Type Description
SOLUTION_TYPE_UNSPECIFIED Default value.
SOLUTION_TYPE_RECOMMENDATION Used for Recommendations AI.
SOLUTION_TYPE_SEARCH Used for Discovery Search.
SOLUTION_TYPE_CHAT Used for use cases related to the Generative AI agent.
SOLUTION_TYPE_GENERATIVE_CHAT Used for use cases related to the Generative Chat agent. It’s used for Generative chat engine only, the associated data stores must enrolled with SOLUTION_TYPE_CHAT solution.

References

Terraform Docs - site://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/discovery_engine_search_engine

Note: I used site://… due to linking limitations for new users.