Repeatability
High
The same six fields need to be extracted from every document using consistent logic. While document formatting varies, the extraction rules themselves don't change per client — this is structurally repetitive and well-suited to automation.
Ambiguity Tolerance
High
Success criteria are concrete: six named fields, standardized date format, 90-day expiration flag, CSV sorted by urgency. A non-human can verify completeness and flag missing values without subjective judgment.
Data & Tool Availability
Medium
PDF files and email attachments must be accessible to the agent — this requires file system or inbox access and reliable OCR. Emails add unstructured variability that can trip up extraction if threading or attachment formats are inconsistent.
Error Cost
High
A wrong expiration date or missed policy could cause a broker to miss a renewal window, damaging a client relationship or creating a coverage gap. Errors are reversible if caught, but the downstream business risk is real and warrants human review of the output.
Human Judgment Required
Low
No taste, ethics, or relationship context is needed to extract and sort structured fields. Edge cases like ambiguous dates or duplicate records benefit from human review, but the core task doesn't require human intuition.