A global leader in industrial equipment and services set out to modernize a critical Extended Warranty platform that underpins its global operations. The system manages warranty coverage, claims, and lifecycle events for equipment used across multiple regions and industries, making it a strategic foundation for future digital innovation.
Challenge: Retire a high‑risk legacy mainframe without slowing the business
The organization’s Extended Warranty workload was running on a high‑maintenance legacy mainframe, tightly coupled to Db2 schemas and decades of embedded business rules. Slow change cycles, dependence on scarce mainframe skills, and rising operational risk made it difficult to evolve the platform in line with the company’s digital ambitions.
Key challenges included:
High maintenance and operational risk from an aging mainframe and complex batch workloads
Slow change cycles made it hard to introduce new warranty products and experiences
Tight coupling to legacy Db2 schemas, increased the cost and risk of every change
Limited observability and automation constrain confidence in large-scale modernization
Solution: Behavior‑first modernization using AWS, PostgreSQL, and AI
To deliver a focused modernization of the Extended Warranty workload, Thoughtworks partnered with a specialist mainframe modernization provider, Mechanical Orchard. Thoughtworks leveraged Mechanical Orchard’s Imogen platform as a complementary capability within its broader delivery framework, supporting behavior-first analysis and the safe modernization of critical legacy components while maintaining functional equivalence.
This approach combined cloud-native architecture on AWS and AI-driven automation to reduce risk and compress timelines, ensuring a seamless modernization experience aligned with AWS best practices.
The approach enabled us to dramatically accelerate modernization timelines. What was originally scoped as an 18-month effort was delivered in approximately five months.
The team targeted the most critical components of the workload and re‑platformed them to a modern stack:
Migrated four key batch jobs from mainframe JCL to Python running on AWS Batch, preserving existing business behavior while shedding mainframe constraints.
Moved three essential Db2 schemas to PostgreSQL, establishing a scalable, cloud‑native data foundation aligned with the organization’s long‑term architecture roadmap.
Applied an AI‑driven, behavior‑first modernization platform to analyze the legacy codebase, understand current behavior, and guide refactoring — ensuring functional equivalence as workloads moved off the mainframe.
- Implemented an automated data validation framework to continuously compare behavior between the legacy and modernized systems, reducing migration risk and increasing confidence in each cutover.
- Used Generative AI–assisted tools to accelerate testing and refactoring, helping engineers analyze legacy code paths, generate and refine tests, and streamline the transition of Java and SQL assets.
This approach enabled incremental modernization, with each step validated against real production behavior rather than static documentation. Feature toggles and automated validation supported low‑risk cutovers, ensuring warranty operations could continue uninterrupted as components moved from the mainframe to AWS.
The resulting architecture for the Extended Warranty workload is built on a modern technology stack, including:
AWS Batch for orchestrating and running batch workloads
Python for migrated batch logic
PostgreSQL as the target relational data store
- Java‑based services integrated with the new data layer
- GenAI and AI‑assisted tooling to support ongoing analysis, refactoring, and validation
Outcome: Faster modernization, reduced cost, and a repeatable pattern
The engagement showed that even complex, business‑critical mainframe workloads can be retired quickly and safely when paired with behavior‑first, AI‑assisted modernization and strong automation.
Highlights included
An 80% acceleration over initial timeline estimates for migrating targeted workloads
Successful migration of four batch jobs and three schemas, with hundreds of Java classes and over a thousand SQL queries transitioned into production
Cost reduction through the decommissioning of legacy mainframe components associated with the Extended Warranty workload
Improved time‑to‑market, with product and engineering teams able to focus more on new features and less on maintaining legacy infrastructure
A repeatable modernization pattern combining behavior‑first analysis, AI‑assisted tooling, automated validation, and feature‑toggled cutovers that the organization can now apply to additional mainframe workloads
About the client
The client is a global industrial equipment and services company, recognized for its engineering excellence and long‑term customer relationships. Its products and services support customers in sectors such as agriculture, construction, and infrastructure, and its digital platforms are central to delivering reliable, high‑quality lifecycle support worldwide.
