Inbound Document Intake
Turn any inbound document into a structured record.
Receives whatever artefacts external parties send — PDFs, emails, scans, EDI — and extracts the structured fields your process needs, without requiring senders to change their behaviour. Low-confidence extractions queue for human review rather than silently corrupting downstream data.
Shape
Operational dimensions
Each output waits on a human decision.
Fires when an upstream condition occurs.
Light state; replaceable any time.
Consumes external data; does not write back.
Inputs
- Inbound artefact (PDF, email, scan, EDI, image)
- Document-type schema defining extraction targets
- Confidence threshold for exception routing
- Inbound channel config (email inbox, upload endpoint, fax connector)
Outputs
- Structured record matching the target schema
- Source artefact reference with provenance
- Exception queue items with extraction confidence trace
- Per-field extraction confidence scores
Mechanism
Receives discrete inbound artefacts (PDFs, emails, scans, faxes, EDI, forms) from external senders, extracts structured fields under a target schema, and lands a structured record plus exception queue for unparseable items.
Why this is a primitive
Cannot be decomposed — receive-parse-extract on inbound artefacts is one operation. Strip the extraction layer and you have an inbox (no primitive); strip the inbound channel and you have a parser (no primitive). The two are inseparable as a single intake mechanism. It does NOT capture live conversation (interview-capture), does NOT host the submitter UI (structured-form-intake), does NOT classify/organise post-landing (organise-* primitives).
Where it shows up
Related primitives
Tags
See where it fits.
Primitives are configured into named solution shapes for each client’s domain. The fastest next step is a conversation about which shape fits your problem.
Start a conversation