Introduction to MDM
Ataccama Master Data Management (MDM) is an advanced, scalable, and highly available platform designed to consolidate and provide master data management tasks.
Ataccama MDM was developed as a fully metadata (data model) driven data processing system, encompassing a metadata editor, a data processing engine, and a ONE plan executor, as well as inheriting all ONE Desktop functionality. In other words, MDM provides Master Data Management functions in combination with a strong emphasis on Data Quality as a core part of the process and provides all Data Quality Management functionality through its natively integrated data quality engine.
MDM is typically considered a standalone engine (server) for Master Data Management. Moreover, it also includes a web application that complements MDM by exposing master data through a web interface, with a rich feature set for data governance: browsing, searching, viewing, creating, and modifying data, and issue resolution support.
MDM can also be used in conjunction with a set of additional Ataccama and even third-party tools and artifacts used for administration and management of all other tasks related to MDM-hosted data. Those appear as interfaces for administration (for example, MDM-native Admin Center), data quality issue and exception resolution, data interaction (browsing, editing, authoring, and more), or middleware systems for generic DI, EAI, and BPM tasks.
Business role
-
Serves as a system of reference (ideally in the role of the central MDM hub), delivers the single version of the truth for all mastered data.
-
Enables peripheral data (connected to master data domains) integration and processing (including standardization, cleansing, and enrichment).
-
Centralizes cleansing, standardizing, matching, and merging rules across the data processing line (from source to master) consistently.
-
Processes data in both full and incremental modes.
-
Supports batch and online (both synchronous and asynchronous) processing modes and coexistence of the two.
-
Provides data on demand in both full and incremental modes via online and/or batch interfaces.
-
Prevents duplicates and low quality data from entering it into any connected source system or the MDM hub itself.
Business cases
Selected business cases where solution is based on MDM:
-
Customer Data Integration (single domain MDM, see Solution scope)
-
Address Data Integration (single domain MDM)
-
Vehicle Data Integration (single domain MDM)
-
Blocklist Data Integration (multiple domain MDM)
-
Group-wide Customer and Product Data Consolidation (multi-domain MDM)
-
Regulatory Compliance Data Integration and Reporting (multi-domain MDM)
-
Analytical and Operational CRM Support (multi-domain MDM)
Solution scope
-
Single domain: MDM maintains one relatively isolated master data domain.
-
Multi-domain: MDM maintains master data domains in some relation, for example, mastered data domains are processed together with their mutual relationships to keep a certain (logical) level of referential integrity.
-
Multiple domain: MDM maintains master data domains that are unrelated (for example, a list of blocklisted addresses and persons, but addresses do not relate to persons).
The individual domains are processed independently although they are maintained within and provided consistently from a single integrated MDM platform. Domains could be interconnected subsequently in business processes (BPM, business process workflow systems) or on wider data integration platforms, for example, ESB or ODS.
Implementation styles
-
Consolidation style: Consolidates a number of systems to a single data structure, identifies possible duplicates, and provides consolidated master records representing groups of related source data records into downstream data integration systems.
-
Centralized style: Master data exists in MDM data hub only (R/W).
-
Mixed style: A combination of Centralized, Coexistence, and Consolidation styles, where master data exists both in the MDM hub (R/W) and source systems (R/W).
This means that if a company has consolidated all data sources but some information is missing, the user is capable of, for example, creating a missing party record with all related entities or adding a missing related record such as address, contact, or consent.
-
Coexistence style: Provides consolidation of source system data, includes upstream propagation of consolidated master data back to the source systems, with data authoring remaining in the originating (source) systems.
Was this page useful?