Who can use this feature?

• Odie admins or users who can manage integrations
• Availability depends on your Odie plan and configuration

Connect Odie to Amazon S3 to import CSV files on a schedule. Create one reusable S3 connection for your organization, then use that connection across scheduled imports.

Understand the S3 import setup

Amazon S3 is the supported source for this release. SFTP support is planned for a future release.

Prepare your S3 access

Before opening the connection wizard, prepare:

1. An S3 bucket that contains the CSV files you want to import.
2. The bucket's AWS region, such as us-east-1.
3. A dedicated IAM user with least-privilege access to the bucket and prefix.
4. The IAM user's access key ID and secret access key.

Create the S3 connection

1. In Odie, open Settings > Integrations > S3 Scheduled Import.
2. Click Connect source.
3. In the Source step, select Amazon S3.

4. In the Credentials step, enter the connection details:

   Field	Required	Description
   Connection name	Yes	A recognizable name for the connection.
   Access key	Yes	The access key ID for the dedicated IAM user.
   Secret	Yes	The secret access key associated with the access key.
   Region	Yes	The AWS region that contains the bucket.
   Bucket	Yes	The bucket name without an s3:// prefix.
   Base folder prefix	No	The root folder within the bucket that all imports using this connection should start from.

5. Click Test connection. You must pass the test before you can save the connection.
6. In the Review step, confirm the details and click Save connection.

The saved connection appears in the connections list and can now be selected when you create a scheduled import.

Grant the required AWS permissions

The IAM user must be able to list the configured bucket or prefix and read objects from it:

• s3:ListBucket on the bucket, scoped to the prefix whenever possible.
• s3:GetObject on objects under the prefix.

Replace YOUR-BUCKET and your/base/prefix in this example policy:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "ListBucketWithPrefix",
      "Effect": "Allow",
      "Action": ["s3:ListBucket"],
      "Resource": "arn:aws:s3:::YOUR-BUCKET",
      "Condition": {
        "StringLike": {
          "s3:prefix": ["your/base/prefix/*"]
        }
      }
    },
    {
      "Sid": "ReadObjectsUnderPrefix",
      "Effect": "Allow",
      "Action": ["s3:GetObject"],
      "Resource": "arn:aws:s3:::YOUR-BUCKET/your/base/prefix/*"
    }
  ]
}

If you do not configure a base folder prefix, grant access at the bucket level. This is broader access, so use a prefix whenever possible.

Create a scheduled import

After saving the connection, create a scheduled import as a separate step:

1. Choose the saved S3 connection.
2. Optionally enter a folder override. Odie adds this path after the connection's base folder prefix.
3. Enter a file pattern, such as accounts_*.csv, if you want to import only matching files.
4. Choose the object type and configure the import schedule.
5. Review and save the scheduled import.

Understand how folders combine

Odie combines the connection's base folder prefix with the optional folder override on the scheduled import.

File pattern, object type, schedule, and the folder override belong to the scheduled import—not the reusable S3 connection.

Understand the connection test

Manage connection security

• Access keys are stored encrypted and are not shown again after you save the connection.
• Use a dedicated IAM user and the narrowest practical bucket and prefix permissions.
• Rotate access keys according to your organization's security policy.
• Connection details are read-only in this release. To change credentials, create a new connection or contact your administrator if an edit flow is available in your environment.

Delete a connection

You cannot delete an S3 connection while one or more scheduled imports reference it. Delete or reassign those scheduled imports first.

Deleting a scheduled import does not delete the reusable S3 connection.

Troubleshoot the connection

FAQs

Question: Is the S3 connection the same as a scheduled import?

Answer: No. The connection stores reusable access to the S3 bucket. A scheduled import selects that connection and defines the folder override, file pattern, object type, and schedule.

Question: Where do I configure CSV field mapping?

Answer: Configure field mapping on the CSV connector for the object type. The mapping is shared and does not belong to an individual scheduled import.

Question: Is SFTP supported?

Answer: Not in this release. Amazon S3 is the available source; SFTP support is planned for a future release.