Automate customer onboarding with an agent
The problem
Onboarding workflows in fintech and professional services stall when customers upload mixed, incomplete ID sets.
Traditional KYC APIs fail on complex multi-document packages or require heavy human review when non-standard ID combinations arrive.
How Stipple helps
Feed all submitted IDs into a deterministic scoring engine: checks AFP 100-point primary and secondary rules, validates DOB/name consistency, and reports the exact missing points if incomplete.
Agents can automatically email the user with the exact remaining document needed (e.g. "We need 25 more points — please upload a utility bill").
The call
View Full Recipe →curl -X POST "https://www.stipple.sh/v1/identity-check?scheme=afp_100_point" \
-F "files=@passport.pdf" -F "files=@licence.pdf"
# → { "established": true, "points": 110, "target": 100, ... }Achieve 90%+ straight-through customer verification with instant, helpful feedback on deficient document packs.
How the corpus was built, what was caught, and what the method does not prove.
Questions
- What is the AFP 100-point check?
- The Australian identity-document points system. The endpoint classifies each document in the pack and tallies points deterministically against the scheme, so the arithmetic is reproducible rather than judgemental.
- Does this replace a KYC or AML provider?
- No. It verifies the document pack and the points tally; sanctions and adverse-media screening is a separate check that pairs with it. Treat this as the document layer of onboarding, not the whole obligation.