
Cloud Solutions

-
Source data load – Source data from business applications is copied to Azure Data Lake, where it is initially stored for further transformation and use in downstream analytics. Source data can generally be classified into one of three categories:
-
Structured master data – The information that describes customers, products and locations. Master data is low volume, high complexity, and changes slowly over time, is often the data that organizations struggle the most with data quality.
-
Structured transactional data – Business events that occur at a specific point in time, such as an order, invoice, or interaction. Transactional data is typically high volume, low complexity, and static (does not change over time).
-
Unstructured data – Can include documents, images, videos, social media content, audio, and so on.
-
-
Source master data load – Master data from source business applications is loaded into the MDM application. Source data should be loaded “as is”, with complete lineage information and minimal transformations.
-
Automated MDM processing – The MDM solution uses automated processes to standardize, verify, and enrich data (or example, verify and standardize address data), identify data quality issues, group duplicate records (or example, duplicate customers), and generate master records (also known as golden records).
-
Data stewardship – As necessary, data stewards can review and manage groups of matched records, create/manage data relationships, fill in missing information, and resolve data quality issues. Multiple alternate hierarchical roll-ups can be managed as required (for example, product hierarchies).
-
Managed master data load – High-quality master data flows into downstream analytics solutions. This process is again simplified because data integrations no longer require any data quality transformations.
-
Transactional and unstructured data load – Transactional and unstructured data is loaded into the downstream analytics solution, where it is combined with high-quality master data.
-
Visualization and analysis – Data is modeled and made available to business users for analysis. High-quality master data eliminates common data quality issues, and improved insights are gained.​
​