Axtria Insights

Axtria Insights

CASE STUDY
Analysis of Complex Data Model Relationship with Enhanced Visualization

Situation

  • Axtria was servicing a banking Client with global presence
  • Geographies across locations were using multiple financial models across multiple data warehouses and variables

Challenges

  • Client was using conventional relational  database management system for managing models across portfolios, BUs and geographies
  • There was interrelationship among models and a lack of efficient ways of predicting impact of a variable or model on other models and scenarios
  • Model validation was a challenge as it was difficult to forecast how a failure or success of a model will impact performance of other models
  • Query based information extraction with minimal visualization

Approach

  • Axtria used Graph database for managing the data and relationship
  • Relationship technique links were created using graph database's node
  • Key nodes were defined – “Model”, “Variable”, “Table”, “Data warehouse”,  “Model Owner”, “Business Owner” and “Business Developer” were defined with relevant attributes and relationship type was defined between various nodes
  • More nodes like compliance could also be added to existing network to enhance models

Result

  • User was able to identify interdependencies without writing complex queries
  • Improved model validation as impact of any node on other node could be viewed easily
  • User was able to view relationship from very high level to the lowest level of hierarchy in the UI itself without writing complex queries

Analysis Of Complex Data Model Relationship With Visualization