One workspace, the whole MLRO workflow

Investigate

Ask anything about AML obligations — filing thresholds, CDD for PEPs, what constitutes structuring — in plain language. ARGOS retrieves the relevant statutes and guidance and streams a cited answer, with the exact sources listed alongside.

• Streaming, conversational answers
• Citations to official legislation & guidance
• Pinnable chat history per session

Screen a name

Type a single name and get an instant verdict against PEP and sanctions data. Optional lanes add an LLM PEP adjudicator (for names the database is unsure about) and adverse-media checks.

• Combined PEP + sanctions verdict
• AI adjudication for data-gap cases
• Adverse-media lane (optional)

Case investigations

Run a full work-up: an entity and its individuals are screened, checked against your internal watchlist, researched against the regulatory corpus, and synthesised into a structured report with an STR recommendation — saved as a case.

• Screening → watchlist → RAG → report
• STR recommendation
• Saved with full history

Source Lists

See exactly which official lists you screen against — sanctions authorities, the PEP base, validators and adverse-media sources — each with its authority, licence, commercial-safe status and live record count.

• Every list, grouped by category
• Licence + commercial-safe status
• Live record counts & attributions

Cases & watchlist

Track open cases through Open → Escalated → Cleared → Closed, and maintain an internal watchlist with fuzzy name matching so near-misses surface during screening.

• Case status lifecycle
• Internal fuzzy-matched watchlist
• Filter, search, resume

Audit trail

Every investigation, ingestion and case action is written to a tamper-evident, SHA-256 hash-chained log. Verify the chain at any time and export it when a regulator asks how a decision was reached.

• Hash-chained, tamper-evident
• Verify integrity on demand
• Exportable evidence

Model registry

ARGOS is not locked to one model. A runtime registry runs per-role, cross-provider fallback chains, fails over automatically if a model errors or is withdrawn, and surfaces newly-available models for your approval — never auto-adopted.

• Cross-provider fallback chains
• Auto-failover & auto-downgrade
• Human-gated model upgrades

Multi-tenant

Each firm is an isolated tenant with invite-only access. The regulatory corpus is shared, read-only public law; every case, watchlist, audit row and session is scoped to your tenant.

• Per-tenant data isolation
• Invite-only team access
• Roles: admin / viewer / super-admin