I’m looking for 1 narrow workflow case for Phase 1 testing.
I’ve already tested a few workflow lines around boundary-oriented behavior.
The next thing I want is one case that gets closer to semantic correctness.
I’m specifically looking for:
- an invoice-document classification workflow
- a small explicit label set
- at least one example where the expected label is already known
- ideally one ambiguous example
A strong fit would be something like:
- Invoice
- PurchaseOrder
- CreditMemo
What I want to understand is:
- did the output stay inside the allowed class set?
- did it match the expected label for this specific case?
- what breaks downstream if the label is wrong?
- how should ambiguous cases be handled: hold, review, or fail_safe?
A short outline is enough first.
I do not need a full production payload immediately.
Helpful details:
- rough payload or document shape
- target schema
- label set
- one example with an explicit expected label
- optionally one ambiguous example
- a short note on downstream risk if the label is wrong
- webhook or polling preference
If public sharing is awkward, even a simplified anonymized outline is enough.