NEXUS
All agents
06Capital Markets & Risk

AURA NEXUS

Fraud detection, credit scoring and structured AML/KYC compliance.

8
sub-agents
7
native connectors
4
key capabilities
The legacy data challenge

Financial institutions combine Murex, Calypso, Summit and dozens of internal risk engines. Decision traceability for credit and AML/KYC compliance is now mandated by DORA, FRTB and the EU AI Act — often without a unified operational framework.

AURA added value

NEXUS argues every credit, fraud or AML decision with its full data lineage, compliant with EU AI Act Art. 14. Traceability is no longer a cost — it becomes an operational advantage.

Description

What NEXUS does for you.

NEXUS detects transactional fraud patterns, scores credit risk in real time and feeds AML/KYC reviews with explainable reasoning. Every alert is argued, reproducible and defensible to regulators.

Targeted read: only the data relevant, impactful or critical to this business enters the ontology. No full IS replication.

Primary connectors
MurexCalypsoSummitBloombergRefinitivWorld-CheckFATF Watchlists
NEXUS — Capital Markets & Risk
Sector illustration · Capital Markets & Risk
Reasoning in progress · Capital Markets & Risk
ingest :: Murex
ontology :: Business · Capital Markets & Risk
argued recommendation
Key capabilities

What you activate
from day one.

01
Real-time transactional fraud detection
02
Explainable credit scoring (EU AI Act Art. 14)
03
Assisted, traceable AML/KYC review
04
Argued FRTB / DORA stress testing
Sector-agnostic postulates

Illustrated on Capital Markets & Risk.
True beyond.

The examples above rely on the AURA NEXUS sector. The principles below, however, hold for any organisation steering a real operation — regardless of the business.

01 · Ontology before model

No generative AI reasons correctly without an explicit representation of business concepts. The model comes after meaning.

02 · Read without migration

Legacy stays in place. AURA reads, never writes, never moves data. Switching cost is zero, full reversibility.

03 · Decision traced, not generated

Every recommendation is tied to the source attributes that produced it. Reproducible. Contestable. Auditable.

04 · Trade-off, not local optimisation

Value arises at the intersection of constraints — finance, ops, compliance, customer. AURA arbitrates; it does not optimise one silo.