
Kumaran delivered a GenAI-powered migration solution built around a tailored LLM pipeline and a human-in-the-loop system to automate code conversion at scale. A custom LLM pipeline was implemented using the Semantic Kernel framework with pre-defined common functions, capable of handling scripts with 6,000+ lines of code. The solution also supported inline SQL procedures, referenced SQL procedures, and dependent KSH scripts for seamless end-to-end migration.
A Tier-1 Global Bank partnered with Kumaran to modernize over 600 legacy KornShell and Sybase SQL scripts against a tight FIS sunset deadline, completing the full migration in just 4 months with over 90% accuracy. The engagement leveraged a custom GenAI-powered LLM pipeline built on the Semantic Kernel framework, with a human-in-the-loop review layer to ensure quality at scale. What would traditionally have required 2+ years and a large migration team was delivered at 25–30% of that team size — enabling the bank to meet its compliance deadline and achieve cloud readiness without disrupting ongoing business operations.
The bank had accumulated over 600 custom KornShell and Sybase SQL scripts built on top of FIS Mid-Office over many years of operation. When FIS announced the sunsetting of that version, the bank faced a non-negotiable migration deadline to a new platform compatible with Python and Oracle SQL. The scale and complexity of the scripts — some exceeding 6,000 lines of code — made a manual migration approach both time-prohibitive and commercially unviable. A smarter, automation-led approach was the only path to meeting the deadline without significant business risk.
Contact us today to learn more about our GenAI-powered migration approach and the success story behind this engagement with the Tier-1 Global Bank.