Case Study 4
Fueling Innovation with AI-powered Code Migration at a Global Travel Technology Leader
Streamlining legacy transformation and boosting development efficiency through intelligent automation
Transforming legacy code into modern languages presented significant hurdles for the client, a global travel technology leader. Manual efforts and outdated systems introduced inefficiencies and skill dependency.
Complex Legacy Code
Navigating the intricacies of zTPF assembly language proved labor-intensive and prone to errors.
Expertise Gap
A lack of experienced developers created roadblocks in reverse engineering legacy systems.
Outdated Documentation
Insufficient and outdated records added layers of complexity to understanding code structure.
Infosys introduced an AI-driven narrow transformer model to make the code migration process easier and more accurate, while reducing manual work.
AI-generated Documentation
Leveraged the AI model to produce accurate business documentation directly from legacy code, addressing documentation gaps.
Reverse Engineering Transformation
With iterative improvements, the model achieved 70% accuracy in extracting business rules, aiding efficient legacy code understanding.
Forward Engineering Approach
Used extracted business rules to define human-readable design rules, enabling accurate Java code generation even without additional training.
Tailored AI Infrastructure
Deployed the Star Coder framework on Infosys' secure network to ensure a reliable and scalable AI application.
The initiative delivered transformational results, reducing dependency on legacy systems and expediting digital innovation.
Reduced Skill Dependency
AI-enabled automation minimized reliance on legacy-code experts, promoting workforce scalability and independence.
Cost Savings in Reverse Engineering
The reverse engineering model slashed both cost and time, enabling efficient system modernization.
Accelerated Transformation
Deployment of AI incrementally enhanced capabilities for large-scale operations, expediting digital transformation timelines.
Strategically deploying AI for legacy system transformation delivers value only when combined with iterative improvement and contextual domain understanding.