Legacy modernization
Reshaping the economics of software development: Building a future-ready core with Al/works™
Enterprise leaders no longer ask whether AI matters. They ask why it still can't move beyond pilots and into the systems that run the business.
The answer is simple: AI can't reason across systems it can't understand. Most enterprises already have the data and workflows AI needs, but that intelligence sits trapped inside decades-old applications and disconnected data estates. Multi-year modernization programs were built for a world where waiting years for value was acceptable. The AI era doesn't allow that.
Modernizing enterprise systems means more than moving to the cloud. It means turning them into adaptive, AI-ready platforms that keep evolving as business needs, regulations and customer expectations change. Here's how Thoughtworks and AWS partner to get you there, combining two AI-enabled technologies: AI/works™ and AWS Transform.
From migration to context recovery
For years, modernization meant migration: move workloads to the cloud, refactor applications, cut infrastructure costs. That's not enough anymore. Cloud migration alone doesn't make an enterprise AI-ready, and simple transpilation can add to the brittleness that blocks AI adoption in the first place.
Enterprises now need to reimagine monolithic systems into architectures where business context, operational behavior and data can be exposed, governed and evolved — for both humans and AI agents. That means recovering the operational semantics buried inside legacy systems: how workflows operate, how business rules interact, how data moves across domains.
The real enterprise asset isn't the code, or even the data. It's the business intent encoded within them. Once you make that shift, modernization stops being a migration problem and becomes a context recovery problem. The organizations that win in the AI era will be the ones that can continuously extract, govern and regenerate that context as their systems evolve.
How AI/works™ and AWS Transform work together
AWS Transform provides the modernization execution engine: decomposition analysis, migration orchestration, modernization acceleration and agentic workflow automation.
AI/works™ provides the enterprise context layer: operational behavior recovery, governed specifications, reusable modernization intelligence, an enterprise context library and continuous regeneration of systems and architectures as they evolve.
Together, they let you reverse-engineer and forward-engineer legacy systems incrementally, while building the AI-ready foundation for what comes next. Modernized architectures become easier for AI agents to reason over. Business context becomes reusable across applications. Transformation outputs become governed enterprise assets instead of one-time deliverables.
Seeing it in practice: a global manufacturer's mainframe exit
A leading global manufacturer needed to retire the mainframe running its extended warranty platform — the system that manages coverage, claims and lifecycle events for equipment used across multiple regions and industries. The mainframe was tightly coupled to Db2 schemas and decades of embedded business rules. Slow change cycles, scarce mainframe skills and rising operational risk made it hard for the business to keep up.
The company set an aggressive goal: retire the mainframe and move to a modern, cloud-based architecture on AWS, without disrupting warranty operations.
Working with Mechanical Orchard's Imogen platform alongside AI/works™, the team took a behavior-first approach — analyze the legacy codebase, understand how it actually behaves, then refactor and migrate while preserving that behavior. They moved four key batch jobs from mainframe JCL to Python running on AWS Batch, and migrated three Db2 schemas to PostgreSQL. An automated data validation framework continuously compared legacy and modernized system behavior, so each cutover could be verified against real production behavior instead of static documentation.
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 result: an 80% acceleration over the original timeline, covering hundreds of Java classes and over a thousand SQL queries. The mainframe is retired, warranty operations never stopped and the company now has a repeatable pattern it's applying across the rest of its legacy estate.
What this means for you
Transformation measured in months, not years. Incremental value delivered along the way, instead of one delayed cutover. Modernized systems your AI agents can actually reason over, instead of disconnected pilots that never scale into production.
The enterprises that succeed in the AI era won't necessarily be the ones spending the most on AI. They'll be the ones that modernize in a way that preserves enterprise context, lowers the cost of future change and keeps regenerating business capability over time. Legacy transformation isn't an IT project anymore — it's the operational foundation enterprise AI depends on.
| Why we earn this work |
| Thoughtworks holds the AWS Agentic AI Competency as a launch partner — one of the first global SIs to achieve it. We also hold AWS Security and Mainframe Modernization Competencies, plus AWS Industry Competencies for BFSI, Retail, Auto & Manufacturing and Life Sciences. We're one of six partners globally on the APD program with AWS ProServe, an AWS Premier Tier Partner and the AWS Global Partner of the Year for Data and Analytics in 2025. Our three-year Strategic Collaboration Agreement co-invests in joint growth. Behind those credentials sit 600+ AWS-certified Thoughtworkers and 30 years of large-scale engineering practice — including the Strangler Fig pattern, authored by Thoughtworks Chief Scientist Martin Fowler, one of the foundational ideas behind incremental legacy modernization that shows up across the industry. |
Next steps for enterprise leaders
Three days. AWS-funded. No cost for the customer to start. The Thoughtworks AI/works™ x AWS Transform Discovery Workshop maps your AI ambition to a Rebuild / Rewire / Reimagine pathway and delivers a working prototype plan.
