Normalized claim
COBOL lines converted: 53,000 lines
53,000 lines of COBOL converted to Java in approximately 1.5 hours
ZenML (Western Union & Unum) agentic AI for mainframe modernization and claims processing using AWS Transform
Use case typeCode modernizationUpdated Jun 13, 2026
Western Union and Unum partnered with AWS and Accenture/Pega to modernize their mainframe-based legacy systems using AWS Transform, an agentic AI service designed for large-scale migration and modernization. Western Union aimed to modernize its 35-year-old money order platform to support growth targets and improve back-office operations, while Unum sought to streamline Colonial Life claims processing and remove fragmented workflows.
- Organization
- Western Union
- Industry
- Finance
- Location
- United States
- Published
- June 2025
Reported outcomes
COBOL lines converted: Approximately 53,000 lines
Time & speed
Discovery/testing timeline: 1.5 monthsPrior timeline for similar work: 3 monthsUnum migration timeline: 3 monthsManual hours eliminated: 7,000 hours annuallyWindows consolidated: 7 windows
Planned next steps
- The source says the organization aims to achieve Prior estimate for Unum work: 7 years.
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Discovery/testing timeline: 1.5 months
infrastructure setup, code transformation, and testing initiation occurred within 1.5 months
Normalized claim
Prior timeline for similar work: 3 months
compared to the previous 3+ month timeline
Normalized claim
Unum migration timeline: 3 months
achieving a complete COBOL-to-cloud migration with testable applications in 3 months
Normalized claim
Prior estimate for Unum work: 7 years
compared to previous 7-year, $25 million estimates
Normalized claim
Manual hours eliminated: 7,000 hours annually
eliminating 7,000 annual manual hours in claims management
Normalized claim
Windows consolidated: 7 windows
Claims examiners consolidated from 7 different windows to a single unified experience
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Western Union, Unum
- Provider
- AWS
- Maturity
- Unknown
- Linked source
- ZenML
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 3 of 3
- 1Legacy Modernization
- 2Workflow Automation
- 3Agentic AI
- Use AWS Transform agentic AI service to analyze, assess, and transform COBOL/mainframe assets.
- Use Amazon Bedrock for model access, Amazon S3 for code storage, Amazon ECS for runtime deployment, AWS Agent Corps for agent registration/orchestration, and VPC/IAM controls for security and isolation.
- Orchestrate specialized agents from AWS, Accenture industry knowledge agents, and Pega Blueprint/workflow components in composable supervisor and linear patterns.
- For Unum, upload code to S3, run AWS Transform to generate business rules extracts, then use Pega Blueprint to visualize and refine workflows before importing them into a Pega cloud-native solution on AWS.
- Western Union converted about 53,000 lines of COBOL to Java in approximately 1.5 hours and shortened discovery/testing initiation to about 1.5 months versus 3+ months.
- Unum achieved a testable COBOL-to-cloud migration in 3 months versus prior 7-year/$25 million estimates.
- Unum eliminated about 7,000 annual manual hours spent by claims managers on routing and work assignment.
Architecture
The architecture uses AWS Transform as the central modernization service, with Amazon Bedrock providing model access and specialized agent capabilities. Partner agents are built in local environments, deployed on Amazon ECS, and registered through AWS Agent Corps. Orchestration follows supervisor and linear patterns using MCP-based communication. Security is enforced via VPC isolation, fine-grained IAM, logging, and observability. In Unum's flow, AWS Transform extracts business rules from code stored in Amazon S3, and Pega Blueprint converts those extracts into workflows for import into Pega cloud-native applications on AWS.
Sources & evidence1
Evidence: Medium50/100Evidence strength
- Customer explicitly identified
- Independent source available
- Quantified outcome available
- Technical implementation details available
Type: Video Or WebinarPublished: Jun 1, 2025Publisher: ZenMLEvidence: SecondaryConfidence: Medium
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
No published comments yet.