ceven

Aryn

Extracts structured data from unstructured documents and feeds the output into your business apps, then monitors document sets for new uploads to trigger automated analysis.

Try Aryn in Ceven

Ask Ceven anything
Standard

Why use Ceven?

  1. AI native Aryn integration

    • Describe the outcome and Ceven picks the right Aryn 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 Aryn data, across all 9 of its actions.

  2. Managed auth

    • Built in OAuth with automatic token refresh and rotation.
    • One place to manage, scope, and revoke Aryn 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 Aryn, when, and on whose behalf.
    • The agent pauses and asks when Aryn is unclear instead of plowing ahead.

  4. Enterprise grade security

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

Supported tools

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

Create DocSet
Use this when you need to allocate a storage container before adding documents to the platform.
Delete DocSet
Permanently remove a docset and all its contents. Use only after confirming the docset id.
Generate plan
Create a query plan without executing it to review the logical steps for a complex data request.
Get DocSet Metadata
Pull metadata for a specific docset including usage statistics and configuration details.
Get Document by ID
Retrieve a specific document record using both the docset id and the document id.
Get Document Binary
Pull the raw binary content of a document for download or external processing.
List Async Tasks
Check the status of all pending or running asynchronous tasks for the account.
Upload Document
Push a new file into a specific docset for parsing and extraction.
Run Query
Execute a data extraction query against a docset to pull specific fields.
Update DocSet
Modify the settings or metadata of an existing document container.
Search Documents
Find documents within a docset based on specific text or metadata criteria.
Cancel Task
Stop a running asynchronous extraction task to save on processing credits.

12 actions · scroll to see them all

Frequently asked questions

Ceven manages large uploads by utilizing the asynchronous task system. When you trigger a batch upload, the agent creates the necessary docsets and pushes the files, then polls the List Async Tasks endpoint to monitor progress. Instead of timing out, the workflow pauses and resumes once Aryn confirms the parsing is complete. You can set up a notification in Ceven to alert you via Slack or email the moment the full batch is processed and ready for extraction. This ensures that even thousands of pages are handled reliably without manual monitoring of the Aryn dashboard.
Yes. The agent uses the Generate plan tool to show you exactly how Aryn intends to query your documents. This logical plan outlines the steps the AI will take to find the requested data points. You can ask the Ceven agent to explain the plan in plain English or suggest modifications to the query parameters to improve accuracy. Once you are satisfied with the proposed approach, you can give the agent a command to execute the actual query, which prevents wasting processing credits on poorly formulated requests.
When Aryn fails to locate a specific field, it returns a null value or a low confidence score. Ceven is configured to detect these gaps and can trigger a fallback workflow. For example, the agent can move that specific document into a manual review queue or send an email to the document provider requesting a clearer scan. You can define the confidence threshold in your workflow settings so that only high certainty data is pushed to your CRM while everything else is flagged for human verification.
Aryn supports most common document formats including PDF, JPG, and PNG. However, users should be aware that Aryn imposes strict rate limits on the number of concurrent asynchronous tasks depending on your subscription tier. If a Ceven workflow attempts to launch too many simultaneous extraction jobs, Aryn will return a 429 error. Ceven handles this by implementing an exponential backoff strategy, queuing the remaining tasks and processing them sequentially to ensure no data is lost, though this may increase the total time for very large datasets.
For compliance and audit purposes, you often need the original file alongside the extracted data. Ceven uses the Get Document Binary action to pull the original file from Aryn and can simultaneously upload it to your own secure storage like AWS S3 or Google Drive. This creates a permanent link between the structured data in your database and the source document. The agent can automatically name these files using the extracted metadata, making it easy to find the original source of any specific data point during an audit.
Absolutely. You can build a workflow where the agent creates a new Aryn docset based on a trigger, such as a new project being created in your project management tool. The agent calls the Create DocSet tool, names it according to your naming convention, and then begins routing all related project documents into that specific container. This ensures that your data remains organized and that queries are scoped to the correct set of documents, which improves both the speed and the accuracy of the AI extraction process.
Data privacy is handled at the Aryn account level. Generally, Aryn provides enterprise options to ensure that your uploaded documents and the resulting extractions are not used to train their global models. When you connect Aryn to Ceven, we only access the data necessary to execute your workflows. We do not store your documents on our own servers; we simply act as the orchestration layer that moves the data from Aryn to your destination system. You should verify your specific data processing agreement with Aryn to confirm your privacy settings.
Aryn is designed to handle multi page documents by treating the docset as a cohesive unit. When you run a query, the agent can request data that spans across different pages or even different documents within the same set. For instance, you can ask for the total sum of all invoices in a docset. Ceven manages the iteration process, calling the necessary Aryn endpoints to aggregate the data and then performing the final calculation before delivering the result to your requested output destination.

Alternatives to Aryn

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

Rossum logoRossumHyperscience logoHyperscienceInstabase logoInstabase

Try Ceven on your stack

Plug Ceven on top of the tools you already run. Connect Aryn 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