Thought Machine has unveiled an AI solution developed with Amazon Web Services (AWS) Transform to help financial institutions speed up the migration of their legacy core systems.
The solution combines AWS Transform’s tools for examining and reverse-engineering mainframe applications with Thought Machine’s Vault Forge suite.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
It is designed to convert legacy code, including COBOL, into financial products that can be tested in a live sandbox environment.
The process focuses on identifying the underlying business rules rather than translating individual lines of code.
This reduces multi-year processes into just days, highlighted Thought Machine.
It operates within the customer’s secure cloud environment and extracts those rules from legacy applications.
The resulting product logic is written in Python for Thought Machine’s Vault platform.
This separates functions such as interest calculations, fee structures, repayment processes, and account lifecycles from the core infrastructure.
AI agents can therefore convert legacy rules into higher-level code without directly handling the underlying database structures.
AWS Transform first examines legacy applications and organises their business logic into structured Easy Approach to Requirements Syntax (EARS) specifications.
The pipeline then reduces overlapping product variants into a smaller set of modern business specifications.
AWS Financial Services Industries, Worldwide Banking head Charith Mendis said: ” WithThought Machine’s Vault Forge and AWS Transform’s AI capabilities, we enable banks to safely extract business logic from legacy systems, providing a new continuous, automated path from mainframe code to cloud-native core banking – solving one of the hardest banking modernisation challenges for banks and we believe it will fundamentally change how the industry approaches core transformation.”
Vault Forge uses specialised AI agents through Amazon Bedrock to produce Python-based financial products that comply with the Vault software development kit, along with test suites and platform configurations.
Banks retain approval authority through Human-in-the-Loop review stages.
After validation, the resulting financial products can be deployed to a Vault platform sandbox for further assessment.
Thought Machine is also introducing a Core Modernisation Accelerator programme.
Thought Machine CEO Paul Taylor said: “By combining AWS Transform directly with Vault Forge, we are fundamentally rewriting the economics of core modernisation. For the first time, banks can deploy AI to rapidly convert complex legacy mainframe specifications into simulation-tested, cloud-native financial products.”
