.png)
New York Life
Insurance · USA
19% lower call hold time
New York Life builds unified data foundation on AWS, enabling GenAI service assistant and 19% lower wait times
100+ data sources consolidated. Legacy Hadoop replaced. Service Sage GenAI assistant deployed.
Friction
New York Life had 100+ different data sources in an ageing on-premise Hadoop data lake. Data silos led to inconsistent customer experiences and limited AI and analytics capability.
Breakthrough
Building a unified cloud data lakehouse as a foundation, with AI as a first-class citizen from the start, not an afterthought. The same data used for operations, analytics, and AI.
Impact
19% reduction in customer call hold time via Service Sage GenAI assistant. Legacy Hadoop replaced. Unified data platform serving all teams.
Unlock the full analysis with breakthrough, impact, what made it smart and its technical approach below!
Problem
New York Life had 100+ different data sources in an ageing on-premise Hadoop data lake. Data silos led to inconsistent customer experiences and limited possibilities for AI and analytics.
What made it smart
The decision to build a unified data foundation first, before building AI applications. Service Sage only became possible because all data was in one place, accessible under one governance framework.
Technical approach
Amazon Redshift as primary data warehouse, AWS Glue for ETL and data integration, Amazon S3 for storage. HIPAA-compliant secure pipelines. Service Sage: LLM-driven assistant via Amazon Bedrock and SageMaker, giving service staff fast access to relevant policy information.
Strategic lesson
AI in financial services is only as good as the data foundation underneath it. Skipping the data layer does not accelerate AI adoption. It limits it.
Reflection question
How many AI initiatives in your organisation have stalled, not because the AI was not good enough, but because the data underneath was not ready?