Enterprise Analytics Platform
Different understanding of data caused long lead times to deliver reports and new products to market
Efficient analytical reports shortened development and production times and led to fewer errors
A large bank with several development teams, all working independently, each with a different “understanding” of the data and deploying independently generated a difficult deployment situation with multiple opportunities to break some part of the production system.
Business analysts had a difficult time understanding and translating information needs into usable development requirements as there was no consistent description of the data causing long lead times to deliver reports and new products to market.
Data was integrated into a corporate data model in a Unified Semantic Layer. This provided the basis for automating code generation using Ab Initio.
Data quality rules were written using Express>It – the Ab Initio tool that allows business users to not only build rules but also test them right there in the tool before publishing them directly to the production environment.
A continuous deployment pipeline was implemented that supported multiple development teams, with the security of automated testing.
Data quality, monitoring and regulatory reporting are requirements of every bank. This project went further to deliver efficient analytical reporting based on standardized definitions of data, with data quality built reporting directly into the solution. A Data Supermarket containing a catalog with distribution for self-service analytics allows all business analysts to get the data they need in the format they need it.
Due to efficient DevOps within multiple business and development teams the development to production times were reduced with fewer errors as a result of continuous testing in deployment pipelines.
FRAMEWORK & TOOLS
Ab Initio Stack (Continuous>Flows, Express>It, Acquire>It, MDW3, Spec-To-Graph, Query>It, Metadata Hub, Data Quality Environment, Testing Framework)
Hadoop (HDFS, Hive)