ceven

Google Cloud Vision

Extracts text and objects from images, labels products for visual search, and monitors long running vision operations to automate asset tagging and auditing.

Try Google Cloud Vision in Ceven

Ask Ceven anything
Standard

Why use Ceven?

  1. AI native Google Cloud Vision integration

    • Describe the outcome and Ceven picks the right Google Cloud Vision calls, fills the parameters, and checks the result.
    • Structured, agent friendly tool schemas so each call runs reliably instead of by guesswork.
    • Rich coverage for reading, writing, and querying your Google Cloud Vision data, across all 29 of its actions.

  2. Managed auth

    • Built in OAuth with automatic token refresh and rotation.
    • One place to manage, scope, and revoke Google Cloud Vision access.
    • Per user and per environment credentials instead of shared keys.

  3. Agent optimized design

    • Actions are tuned from real success and error rates so reliability climbs over time.
    • Full execution logs so you always know what ran in Google Cloud Vision, when, and on whose behalf.
    • The agent pauses and asks when Google Cloud Vision is unclear instead of plowing ahead.

  4. Enterprise grade security

    • Fine grained access so you control which agents and people can reach Google Cloud Vision.
    • Least privilege by default, read scopes first and only the writes a workflow needs.
    • A full audit trail of every Google Cloud Vision action to support review and sign off.

Supported tools

Every action Ceven's agents can run on Google Cloud Vision, and when to use it.

Create Vision Product
Use this when you need to register a new product in a specific project and location after preparing the product details.
Create Reference Image
Use this when adding a new image to a product for detection purposes.
Delete Product
Use this to permanently remove a product and all its associated reference images from the project.
Get Product
Pull the metadata and details for a specific product using its resource name.
Get Product Set
Pull metadata for an existing product set using its full resource name.
Import Product Sets
Use this to bulk import reference images into product sets using a CSV file hosted in Google Cloud Storage.
List Index Endpoints
Pull a list of all index endpoints in a project and location, including pagination details.
Purge Products
Use this to delete products in a product set or orphan products at scale.
Update Product
Modify the display name, description, or product labels for an existing product resource.
Add Product to Product Set
Link a created product to a specific product set within the same project and location.
Get Vision API Operation
Poll the latest state of a long running vision operation to check if it is complete.
List Products in Product Set
Pull all products associated with a specific product set, with optional pagination.
List Locations
Tool to list available Vision AI service locations for a project. Use when you need to discover supported regions before making region specific API calls.
List Vision API Operations
Tool to list operations that match the specified filter. Use when you need to retrieve all operations under a specific project and location.
Update Product Set
Tool to update a ProductSet resource. Use when you need to modify the displayName of an existing ProductSet.
Cancel Vision Operation
Tool to cancel a long running Vision API operation. Use when you need to abort a pending or in progress operation.
Delete Vision API Operation
Tool to delete a long running Vision API operation. Use after confirming the operation name.
Delete Product Set
Tool to permanently delete a ProductSet. Use after confirming the ProductSet's resource name.
Delete Reference Image
Tool to permanently delete a reference image. Use when you have confirmed the reference image's resource name.
Get Reference Image
Tool to get information associated with a ReferenceImage. Use when you have the full resource name and need its metadata.
List Projects
Tool to list Google Cloud projects accessible by the authenticated user. Use when you need to enumerate available project IDs and resource names before performing further operations.
List Reference Images
Tool to list reference images for a product. Use when you need to retrieve stored reference images under a specified product resource name, with optional pagination.
Remove Product from ProductSet
Tool to remove a Product from a specified ProductSet. Use after creating both resources in the same project/location to unlink a product from its set.

23 actions · scroll to see them all

Frequently asked questions

Many vision tasks like bulk imports are asynchronous and do not return a result immediately. Ceven handles this by capturing the operation name returned by the initial request. The agent then enters a polling loop using the Get Vision API Operation tool, checking the status at regular intervals. Once the operation reaches a terminal state of success or failure, the agent parses the final output and moves to the next step in your workflow. This prevents the workflow from timing out and ensures that you are notified only when the image processing is actually complete.
In Google Cloud Vision, a Product is the logical entity representing the item you want to detect, such as a specific model of a shoe. A Reference Image is an actual image file that teaches the model what that product looks like from different angles or in different lighting. To make a product searchable, you must first create the Product resource and then attach one or more Reference Images to it. Ceven automates this sequence so you can simply provide a folder of images and the agent handles the resource creation and linking.
Yes. Ceven uses the vision API to perform optical character recognition on images. The agent can extract raw text, detect the language of the text, and identify the coordinates of words and paragraphs within the image. You can build a workflow where Ceven scans an invoice image, extracts the total amount and vendor name, and then pushes that structured data directly into your accounting software. This eliminates manual data entry for physical documents that arrive as scans or photos.
The bulk import process requires a CSV file stored in a Google Cloud Storage bucket. This CSV must contain the product IDs and the URIs of the images in GCS. Ceven can be configured to generate this CSV based on your inventory list and then call the Import Product Sets tool. Because this is a heavy operation, it runs in the background. Ceven monitors the process and will alert you if any specific rows in the CSV failed to import due to invalid image formats or permission issues.
One critical quirk of the Google Cloud Vision API is the quota system, which separates requests by feature. For example, the quota for label detection is different from the quota for OCR. If you run a massive bulk operation, you may hit a rate limit, which returns a 429 error. Ceven handles this by implementing exponential backoff, meaning the agent will wait and retry the request automatically. However, for extremely high volume needs, you should request a quota increase through the Google Cloud Console to avoid processing delays.
Ceven does not store your images. It interacts with Google Cloud Vision using the permissions granted to your service account. When the agent triggers an import or a detection task, it passes the GCS URI to the API. Google Cloud Vision accesses the image using the internal IAM permissions of the project. As long as the service account used by Ceven has the storage object viewer role, the images remain secure and encrypted within your own cloud perimeter without ever leaving the Google ecosystem.
Yes. Managing the lifecycle of vision resources is a core part of the integration. You can create a workflow that identifies products that have not been searched for in ninety days and then uses the Purge Products and Delete Product Set tools to remove them. This is important because maintaining thousands of unused reference images can clutter your project and make management difficult. Ceven can automate this cleanup on a monthly schedule to keep your visual search index lean and performant.
Yes. Google Cloud Vision supports multiple locations for its product search features. Ceven allows you to specify the location in each API call. You can use the List Locations tool to discover which regions are available for your project. This is particularly useful for global companies that need to keep their product data in specific geographic regions to comply with local data residency laws. The agent can manage separate product sets across different regions within a single unified workflow.

Alternatives to Google Cloud Vision

Other tools that solve a similar problem. Ceven supports these too, so you can switch or run more than one at once.

Amazon Rekognition logoAmazon RekognitionAzure AI Vision logoAzure AI VisionClarifai logoClarifai

Try Ceven on your stack

Plug Ceven on top of the tools you already run. Connect Google Cloud Vision and the rest of your stack, describe the outcome, and its agents handle the work end to end, days of it in minutes.

Get started for free