Integration · Amazon S3

Connect S3 as a knowledge source

Synchronize documents from Amazon S3 or compatible object storage into a searchable Cloud Vault data space.

S3 is a knowledge source in meinGPT. Selected objects from Amazon S3 or a compatible store synchronize into a Cloud Vault data space so chats and assistants can search approved documents. This path does not write back to the bucket.

An administrator configures the endpoint, bucket, credentials and an optional base path. The first run imports objects within that scope; later runs synchronize changes incrementally. The resulting data space has its own access controls in meinGPT.

Technical frame
Amazon S3
meinGPT
Source
Amazon S3
Capability
Synchronize, index and search documents from object storage
Connection
S3 access key, secret, endpoint and bucket; optionally limited to a base path
Data residency
Cloud Vault source; no dedicated Outpost required

Source, permissions, and data residency are documented separately for each connector.

Setup in the docs

Typical workflows

  • Search document archives
  • Summarize document sets
  • Keep knowledge current

What the integration makes possible

01

Search document archives

Find relevant passages across approved objects without browsing bucket paths manually.

02

Summarize document sets

Combine information from several files into one structured overview.

03

Keep knowledge current

Bring changed objects into the data space on later synchronization runs.

Review the prompt and result openly

Prompt

Summarize the current supplier requirements in the approved S3 prefix and cite the source objects.

Result

The requirements cover certification, delivery documentation and annual audits. Sources are suppliers/standard.pdf and suppliers/audit.md.

Does this connection fit your use case?

We clarify the source, permissions, and connection required for the concrete use case. If the existing integration fits, you can move straight to setup afterwards.

Security and operations

Use a dedicated read-only IAM identity and restrict it to the required bucket and base path. Grant access to the data space separately in meinGPT.

Known limitations

  1. 01The integration indexes supported document content; it does not modify objects in S3.

Frequently asked questions

No. S3 is currently a Cloud Vault source and is configured without a dedicated Outpost.