Source-level evidence
Every signal, score and recommendation links back to the underlying record — the email, the filing, the transaction, the document — with the source visible at a click.
Reduce false positives. Explain every alert.
Compass enriches AML and compliance workflows with relational context, helping teams distinguish real risk from operational noise — with traceable reasoning every analyst can defend.
The problem
Most alerts are noise. The real ones are buried.
Rules-based monitoring generates volume, not clarity. Analysts spend their day clearing false positives while genuine relational risk — shared addresses, repeated counterparties, ownership chains — stays invisible across systems.
How Compass works
Context before disposition.
- 01
Ingest transactions, KYC profiles, sanctions, registries and adverse media
- 02
Resolve entities and link counterparties, beneficial owners and devices into a unified graph
- 03
Fuse alert signals with relational context to promote, declass or escalate
- 04
Generate a source-cited reasoning trail for every recommendation
- 05
Return structured outputs into the case management workflow
Non-obvious relationship
From red flag to network.
Three low-risk alerts on their own. Each transaction passes individual thresholds. Compass connects them through a shared registered address, a common UBO and overlapping legal representatives — promoting the cluster as a single investigable network, with a source-level evidence trail an analyst can confirm, override or annotate.
Outputs & benefits
What you get out.
- Alert prioritization with relational context
- False-positive declassification with traceable reasoning
- Counterparty and UBO network reconstruction
- Evidence chain ready for case management
- Audit-ready explainability for every recommendation
Order of magnitude
−40%
False positives on alert backlog
3×
Faster alert disposition
100%
Recommendations explained with sources
Trust by design
Evidence, explainability, audit — built into every output.
Explainable recommendations
No black-box outputs. Each priority states why it was promoted, which signals contributed, which were declassed, and which model version produced it.
Audit trail by default
Signals, model versions, reviewers and actions are logged and reconstructable later — ready for internal control, regulators and your own governance.
Human-in-the-loop
Compass recommends. Authorized people decide. Every consequential action requires a named reviewer who can confirm, override or annotate the reasoning.
Human-in-the-loop
Compass does not autonomously close, escalate or report alerts. Every recommendation is reviewed by an authorized analyst, with full evidence chain and the ability to confirm, override or annotate the reasoning. Designed to support GDPR and EU AI Act obligations.
See it on your data.
30 minutes. We'll wire Compass to a representative slice of your environment — and walk you through the relationships, signals and actions it would surface today.