Screen
A screening engine identifies a potential match involving a name, entity, or payment party.
Case study / Payment screening
More context before money moves.
AI prepares reasoned OFAC alert dispositions for human review. The same approach points toward an accountable payment screening workflow spanning PEP, adverse media, AML, and fraud.
Loresse has solutioned a data analyst digital twin enabling level one disposition.
The challenge
The alert is fast. The work around it isn't.
A possible match needs more than a binary answer. It needs context, a reasoned disposition, and a person who can stand behind the decision.
In work with global banks and insurers, Loresse focuses on the operational burden around screening: gathering evidence, reviewing exceptions, documenting decisions, and keeping analysts in control. No client is identified here.
The workflow
From potential match to accountable disposition.
Assemble
The agent brings together case context and prior dispositions for a consistent review.
Recommend
It prepares a reasoned clearance or escalation with a written explanation and supporting context.
Review
An analyst can inspect and override the recommendation; the decision trail retains the reasoning.
The value / Per screened record
AI level one: AI disposition accelerates human attention to detail.
~4 sec
AI disposition per screened record
Compared with a five-minute manual-review benchmark.
~75×
Faster per-record processing
Equivalent to approximately 98.7% less processing time before human review.
~0.5¢
AI processing cost per record
Versus $1.25 of analyst labor at $15/hour and five minutes per record.
Per-record timing is rounded from a controlled OFAC screening run using synthetic case data; the five-minute human baseline and $15/hour labor rate come from the original Loresse screening materials. AI cost is the measured run cost divided by records processed. These are processing comparisons, not end-to-end turnaround, realized savings, or production performance. Analyst review, escalations, integration, and operating overhead are not included.
Coverage
Five signals. One reviewable picture.
The disposition solution handles all five screening types — OFAC, AML, PEP, adverse media, and fraud — bringing each through the same reasoned, reviewable workflow.
OFAC
Stop-payment screening alert disposition, with documented reasoning and escalation for potential matches.
AML
The screening and disposition approach can inform wider anti-money-laundering investigation and alert triage; those extensions require separate validation.
PEP
Politically exposed person watch and contextual review, with an evidence pack prepared for the analyst.
Adverse media
Monitoring and contextual assessment of relevant reporting alongside other screening signals.
Fraud
Payment risk signals and exceptions brought into the same reviewable decision workflow.
Governance
The analyst owns the decision.
Analysts review and can override the AI recommendation, with a decision trail that preserves its reasoning. Any increase in autonomy depends on validation against a reference dataset and compliance approval; configuration changes are proposed for review, not applied automatically.
PEP and adverse-media watch, broader AML processes, and fraud review extend the same workflow, customized to your screening policies.