Google Vertex AI + Coeffection
Connect Google Vertex AI so the Coeffection AI Assistant runs Gemini in your own Google Cloud project. Coeffection makes a service account for your workspace, you give that account the Vertex AI User role in your project, and every AI request goes to Vertex AI through it. No API key changes hands: the access is short-lived, limited to Vertex AI, and you end it by removing the role.
Everything you get with Google Vertex AI
4 powerful features, seamlessly integrated
Gemini in Your Google Cloud Project
Once Google Vertex AI is connected, Gemini runs in your own Google Cloud project. Coeffection sends each request to Vertex AI through the service account made for your workspace.
Service Account Access
Coeffection makes a service account for your workspace, and you give only that account the Vertex AI User role in your project. The access it gets is short-lived and limited to Vertex AI, and no other workspace can use it.
Your Region, Your Models
Choose Google's global endpoint or a US region, and the Gemini model for each AI task. On the global endpoint Google may process a request in any location. A model has to be available in your project and region before Coeffection can use it.
Test and Revoke
Test the connection from the AI Assistant page whenever you like. To end access, disconnect in Coeffection or remove the role in Google Cloud.
Setup in minutes
Connect your own cloud account — no key to paste
- 1
In Coeffection, open Admin, Integrations, AI Assistant and set up Google Vertex AI. Enter your Google Cloud project ID and choose a region.
- 2
Copy the command shown there and run it in Cloud Shell. It turns on the Vertex AI API in your project and gives the service account Coeffection made for your workspace the Vertex AI User role.
- 3
Confirm that your organization has a Business Associate Agreement with Google Cloud, then save the connection.
- 4
Select Test connection. Coeffection asks Gemini in your project and region for a short reply to a fixed test prompt that holds no customer data.
No Key to Paste
Coeffection uses a Google service account you grant access in your own project. There is no key to paste or store.
No key to paste. Coeffection makes a service account for your workspace, and you give it the Vertex AI User role in your own Google Cloud project. The access is short-lived and limited to Vertex AI, and you end it by removing the role.
More AI integrations
Explore other tools in this category