“I have already uploaded that.”
A reminder can help. It can also reveal that a system does not know the current position. Asking again is particularly frustrating when a document has arrived and is awaiting review.
We developed a targeted reminder workflow for FRED. It checks outstanding requests and the handling status before preparing the next reminder. AI contributed to building and refining the workflow.
Check what is actually missing
The implementation distinguishes open requests from documents already received. Recent work by the team can also make an automatic follow-up inappropriate.
Illustrative example: one document is outstanding while another is already awaiting review. The reminder should identify the open item. This fictional example contains no actual documents.

Targeted follow-up instead of a blanket mailing
| Before the next step | Why it matters |
|---|---|
| Is the request still open? | Do not request completed items again |
| Has a document arrived? | Distinguish receipt from final review |
| Is the team already working on it? | Account for parallel handling |
| Is sending permitted and due? | Follow the configured process |
Sending and its outcome matter too. Preparation, sending rules and logging are separate components in the implementation. Their existence does not establish every current setting.

Automation starts with understanding the process
The main task is not composing a friendly sentence. The system needs to understand an outstanding requirement and which source establishes its status. We connected that check to the available information.
Active reminders are configured for each deployment.
Sources.
- SolvraONE: own development and technical source reviewOwn source · As of



