End-to-End Compliance Platform

Flexible Import

Any format, any structure. Each entry classified automatically.

Upload your screening list in any standard format, or paste names directly. We accept structured data (Excel, CSV, JSON) and unstructured documents (PDF, Word). Each entry is classified as person, organisation, or vessel — that classification drives the matching logic throughout.

What the import handles

Structured import: Excel and CSV with column detection. Name, date of birth, nationality, and other fields are mapped without manual configuration.
Document import: PDF and Word are run through AI entity extraction. Names are picked out of free text and classified.
Direct entry: Type or paste names for ad-hoc screening.
Entity classification: Person, organisation, vessel, or aircraft — set automatically. The class determines matching thresholds and scoring logic.

Structured Review Workflow

Every match reviewed, every decision logged.

Potential matches don't land as a raw list — they run through a structured workflow. Each result is risk-tiered, routed to the right review stage, and needs an explicit decision with a written rationale. Nothing is silently discarded.

The five-stage pipeline

Screening: Names are matched against every active source in parallel. Results above the threshold enter the workflow.
Risk-tiered workbasket: Results sorted into zones by score. High-confidence matches surface first. Low-risk results can be bulk-cleared above a configurable threshold, capped so true positives never slip through.
Validation: Each result reviewed with full context — match score, source, listed reason, legal basis, biographical comparison.
Decision: Accept or dismiss, with a required rationale. Every decision is timestamped and attributed to the reviewing user.
Export: Completed reviews produce an audit-ready report on one click.

AES-256-GCM Encryption

Field-level encryption, no third parties in the screening path.

Screening data is encrypted field by field with AES-256-GCM and HKDF key derivation. The encryption runs on your infrastructure — no data is sent to external services during screening. Counterparty names, match results, and decisions never leave your environment in the clear.

Encryption details

AES-256-GCM: Authenticated encryption that catches both tampering and unauthorised access.
HKDF key derivation: Each field gets its own derived key. No single master key opens everything.
At rest and in transit: Encrypted in the database and over every connection.
No external API calls during screening: The matching workflow runs entirely on your infrastructure. Counterparty names never reach third-party services.
On-premise deployment: The whole platform runs in your environment. Data sovereignty by architecture, not by policy.

Audit-Ready Export

One click, full documentation, ready for the regulator.

When regulators, auditors, or internal compliance teams ask for documentation, it is already there. Each screening session produces a complete report covering every match considered, every decision, every rationale, plus the timestamp and reviewing user for each action.

Documentation falls out of the workflow itself — no separate report-writing pass.

What the export contains

Full decision trail: Every result reviewed, accepted, or dismissed, with rationale and reviewer identity.
Timestamps: Date and time on every action, from screening to final decision.
User attribution: Every decision linked to the reviewing user. The audit trail is individual, not aggregate.
Match context: Score, source, matching algorithm, listed reason, and legal basis for each result.
Project-level summary: Headline numbers for the session — total screened, matched, cleared, accepted.

Alert-Screen & Autoscreen

Ongoing monitoring with automatic re-screening and same-day alerts.

A positive decision enrols the entry for ongoing monitoring automatically. Every night, Autoscreen re-checks every enrolled name against the freshly imported lists. New listings, delistings, or changed identifiers trigger an alert immediately.

How ongoing monitoring works

Automatic enrolment: Confirmed positive matches are enrolled for monitoring without any manual setup.
Nightly re-screening: Every enrolled name is re-screened nightly against the latest data. Changes surface within 24 hours of a list update.
Alert types: Distinct alerts for new listings, delistings, changed identifiers, score changes, and new sources.
Included in every plan: Alert-Screen and Autoscreen ship with every tier. Monitoring is not a paid add-on.

AI-Powered Match Analysis

Plain-language briefings from official sanctions data.

For each potential match, an AI briefing pulls the official record into a structured summary: who is listed, why, under which programme, since when, and what restrictions apply. Reviewers get the context without parsing the legal text by hand.

A chat interface answers follow-up questions against the sanctions data. The AI runs on Anthropic's Claude API — match context is sent over an encrypted connection. Optional, and can be disabled per project.

What AI analysis covers

Listing summary: Programme, designation date, listed reason — in plain language.
Legal basis: Regulation CELEX reference and amendment history, where available.
Biographical comparison: Side-by-side of name variants, dates of birth, nationalities, and identifiers between query and sanctions record.
Follow-up chat: Reviewers can ask specific questions. The AI answers from the official data, not general knowledge.
Data handling: AI runs on Anthropic's Claude API over encrypted TLS. No data retention by the AI provider beyond the request. Disable per project at any time.

PEP Screening

Politically Exposed Persons — 16 official registers covering 14 G20/EU jurisdictions plus Wikidata.

Sanctions screening and PEP screening cover different risks but share the same problem: name matching across transliterations, aliases, and spelling variants. PEP screening uses the same matching engine, pointed at official parliamentary and government registers.

PEP data covers ≈45,900 persons from Wikidata (including Relatives & Close Associates), the CIA World Leaders register, and 14 national parliamentary registers: Bundestag, the 16 German Landtage, EU Parliament, UK Parliament, US Congress, French Sénat, French Assemblée Nationale, Polish Sejm, Norwegian Storting, Danish Folketing, Dutch Tweede Kamer, Swiss Federal Assembly, Austrian Parliament, and the Brazilian Câmara dos Deputados. All sources are integrated into the screening index and refreshed monthly.