Compliance
insights.

Engineering deep dives, regulatory analysis, and product updates from the Veris team. Written for compliance officers, CTOs, and risk professionals.

Latest

New posts weekly

Engineering. Regulation. Product.

Three content pillars covering technical architecture, regulatory developments, and product capabilities.

Engineering

ML model architecture, cross-chain tracing algorithms, and infrastructure scaling decisions explained.

ML models documented

Regulation

GENIUS Act analysis, MiCA implementation guides, and FinCEN advisory breakdowns for compliance teams.

Jurisdictions covered

Product

Feature releases, integration guides, and use case walkthroughs for Veris customers and prospects.

Product modules

Pipeline

Select a stage to see how it works.

Real-time ingestion

Go indexers connect via WebSocket. They capture stablecoin transfers at block confirmation and write to TimescaleDB. Dust below $1 is filtered.

Indexer processes

Sanctions screening

Every address runs against OFAC SDN, threat labels, and mixer registries. Results cache in Redis. Fail-closed: timeouts trigger blocks.

1,332

OFAC addresses

ML risk scoring

Seven detectors and a meta-learner produce a calibrated probability. SHAP explanations and BNN uncertainty bounds accompany each score.

Models

AI triage

The triage agent classifies each alert: false positive, needs investigation, or immediate SAR. Structured output with confidence scores.

2.3s

Triage latency

Investigation

The investigation agent pulls 90-day history, entity profiles, and ZK-circuit analysis. Produces structured evidence with timeline and key findings.

18s

Investigation time

SAR filing

FinCEN Part V narratives with typology codes. A hallucination gate verifies every fact against source data. Transmission via SFTP.

Fabricated facts