SWISS DUE DILIGENCE / GRAPH INTELLIGENCE
See the links.
Understand the risk.
Our proprietary AI/ML engine connects sanctions, PEP, media and corporate records in a time-aware entity graph. Provenance is captured at ingestion, so every relationship can be traced through to an API finding or analyst-prepared dossier.
Follow the ownership path back to the subject, then check the official list entry.
OFFICIAL LIST
01 / THE PROBLEM
A name match
isn't the whole case.
A screening hit may belong to a namesake. A company may be linked through a director or an ownership chain. A headline may lack context. When each result sits in a separate list, the analyst has to reconstruct the network by hand.
02 / THE SOLUTION
Screen the person.
Trace the network.
Screen sanctions, PEPs, media and ownership in the same entity graph. Every finding keeps its subject, relationship path and source close at hand.
Sanctions & PEPs
Explore potential sanctions, politically exposed person and watchlist matches in the context of aliases, connected entities and ownership paths.
WHO IS CONNECTED?Adverse media
Group related reporting across languages, flag syndicated duplicates and link each article to the right subject without losing its source or publication date.
ONE STORY OR SEVERAL?Ownership & control
See how companies, directors and beneficial owners relate, including indirect paths that deserve a closer look.
WHAT IS THE PATH?Temporal monitoring
Compare relationships and source records over time so changes in a subject’s network can be investigated in context.
WHAT CHANGED, AND WHEN?A sanctions match, a news article and a company filing may refer to one person or three different people. GraphLight keeps those records connected to their sources so an analyst can check which is which.
SOURCE LANDSCAPE Swiss SECO sanctions data · other official sanctions lists · public company registers · lawfully accessible media Source availability and permitted use are assessed for each engagement; a public record is not automatically unrestricted for reuse.
03 / THE METHOD
Collect. Connect.
Verify.
Three stages turn fragmented records into a traceable investigation. AI/ML surfaces possible links; source evidence and human review remain essential.
Ingest with provenance.
Source-aware connectors ingest official lists, registries and lawfully accessible records. Each media article keeps its publisher, date, original record and permitted-use context before any grouping.
GRAPH / 01Resolve with AI/ML.
AI/ML-native workflows propose identity matches, group related media coverage and flag near-duplicate syndication. They connect ownership, directorships and signals in our temporal graph, with ambiguity visible for review.
GRAPH / 02Review & deliver.
Follow a finding back to the record and through its history. Apply risk-screening parameters for the mandate, then deliver structured API findings or an analyst-reviewed full dossier.
GRAPH / 03Company and director records connected. Two relationships form the initial view; each keeps its source and effective date.
WATCH / IDENTITY + DIRECTORSHIPFIG. 02 / FOLLOW THE CONNECTION
Risk has
a history.
See it move.
Select a year in the timeline. Ownership adds a new branch in 2024. A directorship ends in 2025, as an affiliate and media record appear. By 2026, the active graph is smaller and a sanctions record is linked to the affiliate. Earlier records remain in the history.
04 / THE GRAPHLIGHT ENGINE
Built from the
source record up.
Ingestion logs, AI/ML entity resolution, temporal relationships and review records share one data trail. Analysts can work back from a finding to the original source.
Provenance at the source
Source-aware connectors and ingestion logs tie each record to its origin, retrieval time, version and permitted-use context. Original articles remain available after grouping.
AI/ML-native entity workflows
Our proprietary engine works across names, aliases and source records to propose identity matches and relationships. Confidence and ambiguous cases are retained for analyst review.
A graph with a clock
Our proprietary temporal graph represents entities, relationships, effective periods and changes. Query connections as they stood on a specific date and compare them with the current graph.
Evidence-led outputs
Keep the relationship path, source trail, risk-screening criteria and analyst review state together for API findings and documented full-report conclusions.
One story. Several sources.
Our media pipeline groups reporting by semantic similarity across languages and flags near-identical syndication. Entity matching connects each article to a potential subject. The cluster shows how a story develops, while the original publisher, date and article remain available for review.
Grouping is not corroboration. A repeated wire story is still one report, even when several outlets publish it.
1 syndication flag
Trace a relationship back to its source.
Provenance is attached at ingestion, not reconstructed when a report is written. Versioned relationships and review logs help teams explain both a finding and its history.
relationship: "owns" subject: "Example subject" object: "Holding AG" valid_from: "2024-03" valid_to: "2026-01" source: "registry record" retrieved_at: "2026-02-10" review: "analyst check required"
05 / TWO WAYS TO WORK
Use the API.
Or request a dossier.
Use structured findings in your own workflow, or ask our analysts to prepare a full report for a specific mandate.
Pay-as-you-go API.
Bring versioned relationships, screening signals and source references into your onboarding or review workflow. Scope each request by relevant risk theme, jurisdiction, relationship depth or time window.
- Structured entities and relationship paths
- Variable screening with related media grouped into threads
- Provenance, effective dates and review state for analyst follow-up
Full diligence dossiers.
Commission an analyst-reviewed report with source-referenced findings, a documented rationale and risk screening tailored to the mandate.
- Identity, corporate and beneficial-ownership review
- Sanctions, PEP, clustered adverse-media and litigation review
- Variable risk screening and risk synthesis mapped to applicable diligence questions, where evidence permits
06 / BUILT FOR REGULATED WORK
Designed to support
defensible diligence.
Swiss financial intermediaries remain responsible for their own legal and supervisory duties. GraphLight supplies evidence and investigation records that can support that work.
Support the questions your file has to answer.
Identity and beneficial ownership, PEP and sanctions exposure, enhanced clarification, ongoing review and documented decisions. Workflows can be scoped around AMLA, applicable FINMA Anti-Money Laundering Ordinance provisions and relevant supervisory or SRO requirements.
Know where a finding came from.
Prioritise official Swiss sources such as SECO sanctions data and relevant public registers, alongside lawfully accessible international and media sources. Record provenance and check access and reuse conditions rather than assuming public means unrestricted.
Protect the process, not just the output.
Handle personal data with the Swiss Federal Act on Data Protection in mind: purpose, proportionality and appropriate safeguards. Flag uncertain AI matches for human review, and separate the evidence from the analyst’s assessment.
Specific controls, source rights and delivery arrangements are defined per engagement. No tool alone makes an institution compliant.
07 / WORK WITH US
See further.
Decide clearer.
Working in Swiss compliance? Let’s scope a pilot for your API workflow or a full diligence dossier, including the sources, review criteria and delivery requirements that matter to your team.
Discuss a pilot info@graphlight.ch