Repeated
the same clearance, every week
A name clears in March and returns in April because nothing recorded why it cleared.
Agentic AI · Screening
Agentic AML-Risk Ops
An agent reads every screening and monitoring alert, decides whether it is a false positive, a true match or one for a human, and then does what your rule says. You write the rule on a canvas.
Minutes
To build a rule
No code
Drag and connect
Dry run
Before anything changes
Every run
Written to the log
Why this matters
The queue is not a workload problem. It is a decision problem.
Most alerts are the same alert. The same name, the same list, the same conclusion, reached again by a different analyst on a different afternoon. What the queue lacks is not people, it is a written-down decision that runs itself.
Uneven
two analysts, two answers
Without a rule the threshold lives in someone’s head, and it moves.
Unwritten
no trail behind the decision
When the regulator asks how the alert was closed, the answer has to be reconstructed.
How it works
Scroll to advance
Four nodes, then the queue works itself
-
Step 01
The alert arrives
A screening or monitoring hit enters the queue with its match rate, list origin and entity type attached.
-
Step 02
The agent forms a view
It returns one of three decisions with a confidence score: false positive, true positive, or escalate.
-
Step 03
Your rule branches
Confidence above 90 closes. Between 70 and 90 goes to an operator with a note. Below that, escalate.
-
Step 04
The action is written down
Close, set match status, assign, tag, note, or fire a webhook. Whatever ran is logged with its result.
sanctionscanner / alerts
- Open alerts1,284
- Auto-eligible912
- Median age2 min
sanctionscanner / agent / decision
- DecisionFalse positive
- Confidence94
- BasisDOB + nationality
sanctionscanner / rule / QNB_Rule
- If confidence> 90
- ThenClose case
- ElseAssign operator
sanctionscanner / agent / log
- ResultApplied
- RuleQNB_Rule
- Retained5 years
Alert
Decide
Branch
Apply
The rule builder
One rule, drawn once
This is the whole rule. A trigger, one condition on the confidence score, and three outcomes. Nothing here is code.
What the builder gives you
A rule you can read out loud
Visual rule canvas
Drag a trigger, a condition and an action onto the canvas. No scripting, no deploy.
43 condition fields
Confidence score, match status, risk level, entity type, amount, channel, article count and more.
11 operators
Equals, greater than, between, in, is true and the rest, so a rule reads like a sentence.
13 actions
Close, set status, set risk, assign, tag, note, notify, convert to case, webhook, toggle monitoring.
Branch and switch
One canvas can hold several outcomes: close the clear ones, route the rest, escalate the few.
Dry run first
Run the rule against live traffic without touching a single case, then read what it would have done.
What it leaves behind
Every run is a record
An agent that acts without a trail is a finding waiting to happen. Each execution writes what it saw, what it decided and what it changed.
Applied
The rule matched and the action was carried out
Logged
Dry run
Simulated only, so you can measure before you switch it on
Logged
Skipped
Matched, but a guard stopped the action
Logged
Failed
The action could not complete, with the reason attached
Logged
FAQ
Before you switch it on
Does the agent act on its own?
Only where you let it. Every action is a node you place yourself, and you can leave a rule in dry run indefinitely.
What if the AI is unsure?
Confidence is a number you can branch on. Most teams auto-close above a high threshold, route the middle band to an operator and escalate the rest.
Can we see what it did?
Every execution writes a log entry: which rule, which node, what it saw, what it changed, and whether it applied, simulated, skipped or failed.
How do we roll it back?
Set the rule to draft and it stops immediately. The log stays, so an auditor can still read what happened while it was on.
Bring a week of your own alerts
We will build one rule on the canvas, run it in dry run against that week, and show you what it would have closed.
Minutes
To first rule
No code
To change one
Dry run
Before it touches a case