ontologyTypes×2
- Customer — name, dob, nationalId (PII, restricted), riskRating
- Alert — alertType, matchScore, sanctionsHit, status
KYC / AML · Reference blueprint
Screen under allowlist. Dispose under dual control.
A reference blueprint for KYC onboarding and AML review — screening and case-disposition agents under strict model and tool allowlists, with every decision explained, dual-controlled, and chained to evidence on the Cortex runtime. This blueprint specifies the ontology, policies, actions, and agents; the signed, installable pack is on the roadmap.
off-allowlist model → 403 · dual-control disposition · reference blueprint (not yet a signed artifact)
kyc-aml-review · Reference blueprint
KYC / AML ReviewBlueprint
ontologyTypes2
policies7
actions3
agents3
reference blueprint · not yet a signed artifact
reference blueprint ▸ ontology · policies · actions · agents
What this pack governs
One governed blueprint — the whole solution, specified.
The KYC/AML blueprint governs customer onboarding and the disposition of AML alerts. It specifies the screening ontology (customer, alert), the policies that enforce allowlists, sanctions-hit holds, and dual-control on adverse decisions, the Action Fabric actions that escalate or close a case, and three curated agents — so a high-scrutiny review workflow is defined as one governed solution that an examiner can take apart.
Representative targets
The targets the pack is tuned to.
Illustrative targets for this use-case, configurable per deployment — the gates that enforce them are specified in the blueprint.
100%
Allowlist enforcement
Off-allowlist model or tool → 403, fail-closed.
100%
Decision explanation
Every disposition carries a cited rationale.
Dual
Adverse-decision control
Representative target — case closure requires two approvers.
What's inside
Four governed sections, one blueprint.
A blueprint specifies the four governed sections a Solution Pack bundles. When it ships as a signed pack, the same envelope — canonicalize → sha256 contentHash → HMAC signature — is what makes it portable and provable.
policies×7
- Off-allowlist model or tool → 403 (fail-closed)
- Sanctions hit → hold, escalate to human reviewer
- Adverse disposition → dual-control (two approvers)
- nationalId is restricted — read denied outside screening
actions×3
- escalateCase — riskTier: medium, opens human review
- closeCase — riskTier: high, dual-control approval
- requestEDD — riskTier: low, enhanced due diligence
agents×3
- KYC Screener — onboarding checks under strict allowlist
- AML Triage — scores alerts, cites the matched signals
- Case Disposer — proposes disposition, never closes alone
What it does
Governed agents, mapped to real Cortex controls.
Every capability in this pack is enforced by a runtime control you can audit — identity, policy, action approvals, and a sealed Trust Ledger. The pack configures them; the platform enforces them.
Strict allowlists, fail-closed
- Agents may use only allowlisted models and tools. An off-allowlist call returns 403 — there is no quiet fallback to an unapproved model on a sanctions decision.
Dual control on closure
- closeCase is a high-risk action requiring two approvers. No single agent — and no single human — can quietly clear an AML alert; the second pair of eyes is policy, not procedure.
Every match is explained
- The AML Triage agent cites the signals behind each match score and sanctions hit. Dispositions carry a rationale a regulator can read — no opaque scoring.
Sanctions hits stop the line
- A sanctions match forces a hold and a human escalation by policy. The agent cannot self-clear a hit; it routes the case to a reviewer with the evidence attached.
Identity data stays restricted
- nationalId and dob are restricted properties — read-denied outside the screening agents that need them, keeping sensitive identity data out of unrelated prompts.
Chained to evidence
- Each escalation and closure is sealed in the hash-chained Trust Ledger with a signed receipt and 10-hop lineage — so a disputed disposition has a tamper-evident record behind it.
403Returned for any off-allowlist model call
100%Dispositions with a cited rationale
2Approvers required to close a case
10-hopProvenance on every disposition
How it ships
A blueprint today — a signed, installable pack on the roadmap.
This blueprint fully specifies the ontology, policies, Action Fabric actions, and agents. When it ships as a Solution Pack it inherits the same packaging every Cortex pack uses: canonicalize → sha256 contentHash → HMAC signature, then a verify-first importPreview → importApply that checks hashOk + signatureOk before a single row is written — into your own tenant only.
- Ontology, policies, and Action Fabric actions specified and ready to package
- On publish: verifyPack checks hashOk + signatureOk before import
- Runs on the same governed runtime as the shipping Tax pack — same gates, same ledger
- Want it sooner? Bring your workflow and we'll prioritize the build.
packs · verify + import
On-allowlist call200 OK
Off-allowlist model403 denied
Sanctions hitheld
Single-approver closeblocked
importApply verifies first ▸ 409 if invalid · upserts into your tenant only
From blueprint to shipping pack
How this blueprint becomes an installable pack.
The design is done; packaging is what remains. A blueprint carries the same four governed sections a shipping pack does — it just isn't sealed and signed yet.
- 01
Specified today
The ontology, policies, Action Fabric actions, and agents above are the real design — the same shape the runtime enforces for every shipping pack.
- 02
Packaged & signed
When it ships, the contents canonicalize → sha256 contentHash → HMAC signature and land in the registry on a stable or beta channel — exactly like the Tax pack.
- 03
Verify & install
importApply re-checks hashOk and signatureOk, then idempotently upserts into your own tenant. Until then, we can stand the workflow up for you from this blueprint.
Governance included
The controls your auditors already cite — built in.
The ontology permissions, fail-closed policies, approval gates, and sealed Trust Ledger map directly to the frameworks this use-case reports against. Aligned with — never claiming a certification you don't hold.
BSA / AML · programOFAC · sanctions screeningFinCEN · recordkeepingSOC 2 · auditEU AI Act · logging
Works with the rest of the platform
The pack is configuration — the runtime does the enforcing.
This pack rides the same governed runtime as every other Cortex solution. Explore the controls it configures and the industry it serves.
Banking & KYC/AML industry
Onboarding, screening, and case-disposition agents under strict allowlists and dual control.
Agent IAM
The model and tool allowlists that make an off-allowlist call return 403 are enforced at issuance.
Trust Ledger
Where every disposition is sealed — verifiable offline when an examiner challenges a decision.
Oversight
Dual-control approval modes for high-risk closures — autonomy can only tighten the gate.
Ship KYC/AML review that an examiner can take apart.
Bring your onboarding and AML workflow and we'll stand up the governed pack from this blueprint — screening under strict allowlists with dual-control on every adverse decision.