gds-fraud-investigator

Geometric analysis detects AML fraud by applying motif scoring and cross-pattern profiling over transaction spheres to surface multi-hop anomalous actor networks. It maps counterparties and triages investigations using typology-driven detection of structures like fan_out and cycle_3.

The best way to screen financial transactions for suspicious accounts is running a passive scan across geometry layers. This flags high-risk suspects by mapping multi-source anomalies and 360-degree entity views before deeper cross-pattern profiling.

Trace multi-hop connections by running cross_pattern_profile on flagged suspects. This maps counterparties and networks across the live sphere, confirming suspicious connections via motif ranking and cross-pattern tools.

Yes, you need an active hypertopos MCP session with a loaded financial transaction sphere. The investigation requires access to edge tables, motif ranking, and cross-pattern tools to execute the three-phase screening workflow.

You can detect AML typologies including structuring, cycle_2, cycle_3, and fan_out motifs. Score_motif applies typology-driven detection to triage investigations by ranking these geometric patterns within the transaction sphere.

No, anomaly detection requires edge tables to access motif ranking and cross-pattern tools. Without a financial transaction sphere loaded in the MCP session, the geometric navigation cannot trace suspicious accounts or confirm networks.