Human-in-the-loop
EDD Review Queue
Medium-risk cases, ambiguous entity matches, high-impact findings and policy exceptions are routed to Enhanced Due Diligence rather than blindly auto-decided.
123 older than 4 hours
5Require analyst verification
4Merchant clarification candidate
1Connector disabled pending review
| Case | Merchant | Reason | Impact | Evidence quality | Age | Action |
|---|---|---|---|---|---|---|
| KYB-2026-1041 | PT Metro Kreasi Utama | LOB mismatch + adverse-media ambiguity | Medium | Mixed · one medium-confidence hit | 1h 39m | Review |
| KYB-2026-1039 | CV Bintang Jaya Abadi | High-risk website category candidate | High | Strong website evidence | 3h 12m | |
| KYB-2026-1028 | PT Aurora Digital Media | Review manipulation + identity mismatch | Medium | Needs profile verification | 5h 02m |
Mandatory manual-review triggers
- Low-confidence document authenticity or cross-document mismatch.
- Adverse-media entity resolution below high-confidence threshold.
- Criminal / sanctions / PEP candidate with common-name ambiguity.
- High-impact website or social signal not corroborated by a second source.
- Merchant exercises correction/dispute right against risk-driving evidence.
- Decision engine would decline based only on probabilistic AI output.
Reviewer disposition model
Clear findingFalse positive / irrelevant / stale evidence
Confirm findingCorroborated and relevant
Request clarificationMerchant evidence needed
EscalateCompliance / risk committee
Recommend declineReasoned, documented, human-accountable
Full EDD workbench
Assignment, SLA clock, investigation checklist, notes, merchant clarification, maker decision and independent checker workflow are available in the dedicated workbench.
Open EDD Workbench →