Thought Machine and AWS Launch AI Core Banking Migration Tool

Updated 1 Oct 2026 · Reviewed by the Zekai editorial team

According to a press release from Thought Machine, the company has launched an AI-powered migration solution in partnership with Amazon Web Services, Inc. (AWS) Transform. The offering aims to help financial institutions modernise legacy systems faster and at lower costs. Rather than executing manual, line-by-line code translation, the joint pipeline is designed to extract business intent directly from millions of lines of mainframe code, such as COBOL, within a secure cloud environment.

The technical architecture pairs AWS Transform's mainframe reverse engineering tooling with Thought Machine's Vault Forge AI tooling suite. The process operates across four distinct stages: extraction, consolidation, synthesis, and validation. During extraction, AWS Transform derives business logic into structured Easy Approach to Requirements Syntax specifications. The consolidation phase then reduces hundreds of redundant legacy product variations into streamlined business specifications.

In the synthesis stage, Vault Forge orchestrates specialised AI agents via Amazon Bedrock to generate SDK-compliant Python financial products, test suites, and platform configurations. The validation stage introduces human-in-the-loop review gates to provide explicit sign-off authority. Once validated, the generated products deploy to a Vault platform sandbox for evaluation, allowing banks to receive migration proofs early in the process without dealing with hardcoded database dependencies.

Traditional legacy cores often hardcode financial products directly into underlying databases, which can make migration projects slow and complex. Thought Machine states that its Vault platform defines financial products entirely as code using Python. Because product logic for interest calculations, fee schedules, and repayment workflows is separated from core infrastructure, the system enables AI agents to translate legacy rules into clean, high-level code.

To support financial institutions adopting the technology, Thought Machine is launching a Core Modernisation Accelerator programme. This initiative places technical architects, product specialists, and forward-deployed engineers directly alongside bank teams. Paul Taylor, CEO of Thought Machine, states that the partnership aims to eliminate technical debt and reduce the time required to complete cloud transitions for large institutions.

Charith Mendis, Head of Worldwide Banking for Financial Services Industries at AWS, claims the solution provides a continuous and automated path from mainframe code to cloud-native core banking. Because the announcements originate from corporate press releases, engineering teams should evaluate these automated pipelines against their own security protocols, code quality requirements, and legacy complexity before committing to production migration timelines.

Why it matters for AI for Finance & Business

For engineering and architecture teams inside financial institutions, this partnership targets the primary bottleneck of core migration: manual mainframe code translation. By utilizing AI agents via Amazon Bedrock to extract business logic into Python-based product definitions, the tooling attempts to reduce multi-year modernisation timelines into days while maintaining human-in-the-loop validation.

What to do about it

• Review the four integrated stages of the migration pipeline to assess compatibility with existing legacy systems.
• Evaluate the Core Modernisation Accelerator programme to determine if external engineering support suits your migration timeline.
• Examine security protocols around running AWS Transform and Vault Forge inside your institution's secure cloud environment.

Source: Thought Machine Partners with AWS to Launch AI-Powered Solution to Accelerate Legacy Core Banking Migrations – Financial IT (financialit.net)