Gemini Enterprise Agent Platform (Vertex AI) prerequisites for the Google Gemini connector in Jitterbit Studio

Introduction

These are the prerequisites for authenticating with the Google Gemini connector using the Vertex AI platform option, which connects to the Gemini Enterprise Agent Platform (previously known as Vertex AI).

Create a service account

Follow these steps to create a service account and grant roles to provide access to the Google Cloud project to be used with the Google Gemini connector:

1. In the Google Cloud Console, create a Google Cloud project or select an existing project.

2. Enable the Agent Platform API for the selected project.

3. In the navigation menu, navigate to IAM & Admin > Service Accounts.

4. Click Create Service Account, located along the top of the page.

5. Enter the Service account details:

   • Service account name: Enter any string for the service account name.

   • Service account ID: This field is automatically populated to match the Service account name and can be edited, if desired.

   • Service account description: Enter a description for the service account.

6. Click Create and Continue.

7. Assign a role to allow access to the Gemini Enterprise Agent Platform (required for a successful connection in the Google Gemini connector):

   • Role: Using the menu, select Agent Platform User (Vertex AI User) as the Role (roles/aiplatform.user) to allow access to the Gemini Enterprise Agent Platform, then click Continue.

8. Optionally, grant access to the service account to specific users.

9. Click Done.

Obtain service account credentials

Follow these steps to obtain the credentials required for connection configuration:

1. On the Service accounts page, click the Actions menu for the service account created above, then click Manage keys.

2. Click Add Key > Create new key, select JSON as the Key type, and then click Create to download the JSON key file to your computer.

3. Open the downloaded JSON file in a text editor and retain the values of project_id, client_email, and private_key to be used during connection configuration.

Enable RAG corpora

To use the RAG corpus feature in the Prompt activity, these additional prerequisites apply:

• The RAG Engine must be enabled for your Google Cloud project.

• One or more RAG corpora must already exist in your Google Cloud project before they can be selected in the Prompt activity configuration. See Create a RAG corpus in the Gemini Enterprise Agent Platform documentation for details.

• The service account must have the RAG Engine User (Vertex AI RAG Engine User) role (roles/aiplatform.ragEngineUser) assigned, in addition to the Agent Platform User role assigned above, to access RAG corpora.