This map shows how columns across Marlink's systems are connected — where vessel IMO numbers, customer IDs, and billing codes flow between systems, and how they relate.
An AI agent analyzed 1,100+ tables — examining schemas, sample data, column semantics, value overlap, and cross-system references — to identify which columns across systems represent the same real-world concept.
Confidence % is the measured value overlap between the two columns — the fraction of the copy's values that actually appear in the source. Items marked for review need your domain expertise to confirm.
How was this built? ▾
How authority (source of truth) was learned — not assumed.
For each family, the agent measured containment across every system: if system A's values are a subset of system B's (A→B near 100%, B→A lower), B is the origin and A copied from it. The system whose values contain all the others is the source of truth. This is corroborated by registry role (e.g. IHS is the external ship registry, authoritative for vessel identity) and completeness. Every edge's confidence is the real value-overlap measured on the full lake — no hardcoded priority table, no name-pattern guessing. Where containment is symmetric or overlap is partial, the edge is flagged for your review with the actual shared values shown.
--
systems mapped
from IHS ship registry to satellite providers, billing, CRM, and more
--
column relationships confirmed
schema connections with production overlap evidence
--
need your expertise
relationships the agent can't confirm without domain knowledge
Relationship types:
Identitysame field, direct referenceCopyindependent duplicateTransformreformatted/mappedDerivedcomputed from source|Confirmed (≥85%)UncertainWeak
Data Lineage
Marlink Maritime Telecom