What it does
The document intake agent watches an inbox or upload folder, reads each document whatever its layout, extracts the fields your process needs and creates a draft record for a person to confirm.
It keeps the source page beside every extracted value, so reviewers check in seconds rather than re-reading the document.
For teams that key the same details from PDFs, faxes and emails all day. The agent drafts the record; a person confirms it.
Uncertain fields are flagged, never guessed.
What it may and may not do
| Operation | System | Access | Note |
|---|---|---|---|
| Read incoming documents | Inbox or folder | READ | |
| Extract fields | Internal | COMPUTE | Confidence per field |
| Suggest a queue | Internal | COMPUTE | Suggestion only |
| Create draft record | Target system | DRAFT | Confirmation required |
| Make a decision on the case | Target system | NEVER | Not built |
How autonomy is earned
Level 1 · Shadow
Extracts silently; results compared with what staff keyed.
Risk: none
Level 2 · Assist
Draft records appear for confirmation, with source pages alongside.
Risk: low, every record confirmed
Compatibility
- Inputs
- PDF, scanned fax, email, Word
- Targets
- Patient administration, case management, ERP
Probably not for you if
- Handwritten documents as the main input
- Processes where extraction is not the bottleneck
Questions we get asked
What accuracy should we expect?
It depends on your documents. We measure it on a sample of yours during discovery and report it field by field before you commit to a build.
