Revolutionizing mainframe and legacy modernization using Gen AI based Copilot for reverse and forward engineering for a Global Travel Tech leader
Overview.
The client is a travel technology company facilitating air bookings and connecting airlines with hotels and travel agencies. Their platform provides real-time information on availability and pricing.
The client needed to migrate their code from a legacy system written in Assembler to Open Systems utilizing Java Microservices-based technologies in a cloud environment. The customer initially followed a traditional approach, which involved a business analyst reading existing code to generate product requirements. Subsequently, the process included design and development using Java technologies. However, this manual approach proved to be both labor-intensive and time-consuming. Seeking a smarter solution, the client wanted to adopt a technology-driven reverse engineering/forward engineering automation to achieve efficiency gains and cost-effectiveness.
Coforge, a long-standing trusted partner, has in-depth domain knowledge of all the applications in the scope of the migration. We have been supporting their applications for over 20 years and were able to understand their pain points.
Solution.
Coforge proposed an innovative hybrid approach which is a mix of leveraging transformative capabilities of Gen AI using GitHub Copilot and expert human interventions to reverse engineer from Assembler code and forward engineer to Java/Springboot based microservices architecture
The Impact.
Gen AI-assisted reverse engineering of Assembler code: Using the power of GitHub Copilot, followed by review and updates by the Assembler developer, business requirements are generated in a shorter time with a higher degree of confidence in a structured and consistent format aiding prompt engineering.
Accelerated target design development using template design patterns: Template design patterns help the designers to define Microservices-based application design quickly and ensure consistency.
Automated Java Code Generation: GitHub Copilot streamlines the code generation process in Java (Spring Boot) from PTSD, minimizing manual effort and accelerating development
Functional & Integration Test Script Generation: The overall development time is reduced drastically due to automated code generation and testing.
Effort Reduction: Coforge's approach reduces SDLC time and effort, resulting in an overall savings of ~ 20-30% and significantly improving design consistency and streamlining the development process.
Key highlights
The client’s decision to partner with Coforge for implementing a Gen AI-based solution is a strategic move towards streamlining their code migration journey. Coforge's deep understanding of clients' applications and their expertise in GitHub Copilot integration and requirements creation, positions them as an ideal partner for this critical legacy modernization initiative. By embracing AI-powered code automation, the client can accelerate migration efforts, reduce development costs, and enhance overall project efficiency.