IBM InfoSphere MDM

IBM Master Data Management, the product formerly sold as InfoSphere MDM, is a cloud-native platform that unifies and governs data across domains. It uses machine learning for entity resolution and relationship discovery, offers SaaS, on premises and hybrid deployment, and provides low-code interfaces and open APIs. The InfoSphere name no longer appears on the product page.

Machine learning for entity resolution is the technically substantive claim and it addresses the hardest part of this work. Deciding whether two customer records are the same person is not a rules problem in practice: names are misspelled, addresses change, companies merge and abbreviate, and every deterministic rule that catches real duplicates also merges records that should stay apart. Learning from resolved cases handles the ambiguity that rules cannot express.

Relationship discovery goes further, identifying that two records belong to the same household or corporate group rather than to the same entity. That distinction matters commercially and is invisible to matching alone.

Low-code interfaces reflect who does this work day to day. Data stewards resolving flagged matches are subject matter experts rather than engineers, and a product requiring code for routine adjudication puts the work back on a team that cannot judge it.