NexusGreenData's master data management builds one trusted, golden version of your core entities — customer, citizen, product, supplier — from duplicated records scattered across systems, and writes it back to your data catalog with lineage and certification.
Master data management (MDM) resolves the same real-world entity recorded differently in many systems into a single golden record. NexusGreenData is a generic, metadata-driven engine: entity types are configuration, so the same platform governs any domain rather than a fixed list.
It shares one data catalog with the rest of the platform, so sources, matched records and golden records all live beside your wider data estate, with lineage and ownership visible to the whole data office.
What it does
Combine exact rules with probabilistic matching (Splink on PostgreSQL). Candidates are scored and clustered so near-duplicates are caught, not missed.
Configurable survivorship rules pick the winning value per attribute and build a versioned golden record you can trust and trace.
Uncertain matches are routed to stewards with capability-based access, who clear the queue with a full audit trail.
Stewards compare candidate records side by side, then merge or split with guarded, reversible actions.
Model relationships between entities and roll them up into hierarchies — household, legal group, org structure — rebuilt under guard rails.
Golden records, lineage and certification are published back to the data catalog, so the whole organisation consumes one governed version.
Register your sources, define an entity type and its attributes as configuration, and run a match. The engine standardises values, generates deterministic and probabilistic candidate pairs, clusters them, and applies your survivorship rules to build golden records.
Stewards clear what the engine is unsure about from the inbox, and the certified golden records — with their lineage — are written back to the data catalog for everyone else to use.
| Matching | Deterministic rules plus probabilistic matching (Splink on PostgreSQL) |
|---|---|
| Survivorship | Configurable rules per attribute (e.g. highest-quality source) |
| Stewardship | Inbox, compare and merge, notes, full audit trail |
| Hierarchies | Relationships and guarded hierarchy rebuilds |
| Catalog | Golden records written back to the data catalog with lineage and certification |
| Entity types | Generic and metadata-driven — any domain by configuration |
| Languages | Arabic and English, right-to-left |
Who it is for
One customer or citizen exists differently in a CRM, a billing system and three spreadsheets, and no report agrees. MDM resolves them into one golden record.
Reference and master data management is an NDMO domain. A working MDM capability turns that requirement into a running, auditable practice.
Data stewards need a governed queue, guarded merges and an audit trail — not ad-hoc edits to production systems.
FAQ
Both deterministic rules and probabilistic matching (powered by Splink on PostgreSQL). Candidate pairs from both strategies are combined, scored and clustered, so records that exact rules miss can still be matched on probability.
Configurable survivorship rules decide which source value wins per attribute — for example the highest-quality source or the most recent value. The surviving golden record is versioned, and steward decisions are recorded.
Yes. A stewardship inbox routes uncertain matches to data stewards, who compare records side by side, merge or split them, and leave notes. Actions are governed and auditable.
Yes. Golden records and their lineage are written back to the data catalog as governed, certified assets, so the rest of the organisation consumes one trusted version through the tools it already uses.
Let us show NexusGreenData MDM on your own duplicated records — matching, survivorship and a golden record written back to your catalog.