File repositories: which ones Dattos connects to and how to set them up
A file repository is where your company's systems deposit and store files: statements, legacy system reports, and response files. By connecting the repository to Dattos, processes can fetch input files automatically and deliver exported files directly to their destination, with no manual sending or downloading.
The integration has two parts:
Set up the connection to the repository once, under Credentials. This article explains how.
Link each process to the repository folders using the Source Directory and Output Directory. See Integration with file repositories: source and output directories.
Repositories with native integration
Credential type | When to use | Tutorial |
|---|---|---|
SFTP - Password and SFTP - SSH Key | Your company's or a partner's SFTP server (bank, acquirer, ERP that generates files) | |
FTP | FTP server that does not yet support SFTP | |
Amazon S3 | Bucket in AWS | |
Azure Blob Storage | Container in an Azure storage account | |
Google Cloud Storage | Bucket in Google Cloud |
SharePoint: there is no native integration. If your company needs to fetch or deliver files in SharePoint, contact the Dattos consulting team to discuss a custom integration. See also Custom integrations via REST API and SQL connectors.
What to prepare on your end
Before setting up the connection, coordinate with the IT team responsible for the repository:
A dedicated access account for Dattos, with its own username, key, or service account. This allows your company to revoke access without affecting other systems.
Permission to read, write, move, and delete files in the folders Dattos will use. The platform reads input files, moves each one to control subfolders (
.processing,.processed, and.error), and writes exported files. If it only has read permission, the process will fail.The folders must already exist. The platform lists the folders in the repository; it does not create the source folder for you.
An SFTP or FTP server accessible over the internet. The platform cannot connect to internal network addresses (for example,
10.x.x.x,172.16.x.xto172.31.x.x, or192.168.x.x) or servers that are only available within your company's VPN. If the server only accepts connections from approved addresses, ask Dattos Support which addresses need to be allowlisted in the firewall.
Set up the connection
Click the gear icon in the upper-right corner and open General › Credentials.
Click + New credential.
On the Storage tab, select the repository type.
Enter the Credential name. This is the name that will appear under Storage Integration in the process settings; use a name that identifies the repository, such as "SFTP Bank X - statements".
Fill in the fields for the selected type. Each type's tutorial explains where to find the required information.
Under Privacy profiles, choose who can use the credential. Only users with one of these profiles will see the repository when configuring a process.
Click Save.
Sensitive fields (password, key, service account JSON) are not displayed again after you save them. When editing, leave the field blank to keep the current value.
Check that the connection works
SFTP: after saving, reopen the credential and click Test connection. The test confirms that the server responds at the specified address and port, but does not check the username and password or the key.
Amazon S3, Azure Blob Storage, Google Cloud Storage, and FTP: do not have a Test connection button.
For all types, you can confirm full access by opening the Source Directory for a Turbo AI process, clicking Edit, then Add destination, and selecting the repository under Storage Integration. If the Import Folders list loads, the connection and read permissions are working. If an error message appears, check the details against the tutorial for your repository type.
Maintain the connection
Change the password or key: edit the credential and fill in only the field that changed. All processes using the credential will start using the new value.
Deactivate: the credential will no longer appear for new connections, and processes using it will show Integration unavailable in the directory editor. To edit an inactive credential, reactivate it first.
Delete: you can only delete a credential when no processes are using it. The Processes column in the credentials list shows how many processes are linked to it; when you try to delete it, the platform lists them.