A UK-based retail bank had built its reputation on digital transformation, operational resilience, and delivering seamless customer experiences across every channel. But underneath that experience sat a SOA-based architecture built on TIBCO. As the bank set out on a broader legacy modernization journey, it knew the first step had to be a focused discovery phase to assess just how ready its systems were for automation. Service discovery was highly manual, making it hard to build a clear picture of legacy systems. Inconsistent documentation practices made reuse difficult and slowed collaboration across teams. Developers were spending significant effort just creating test cases, time pulled away from the core implementation work that actually moved modernization forward. And without intelligent tooling, the bank had limited visibility into how services depended on one another and where business logic existed. The bank’s ambition was clear: lay a strong automation foundation that could give developers AI-driven support to modernize with precision and consistency.
To meet that ambition, Ascendion deployed AAVATM, its GenAI-powered platform, purpose-built to accelerate the discovery phase through domain-aware Agentic AI workflows spanning both analysis and development tasks.
A Microservices Analysis Agent parsed microservices to surface dependencies, endpoints, HTTP methods, and request/response payloads alongside core business logic, while a SOA Analysis Agent worked through legacy SOA components to extract database calls, SIBIS wrapper interactions, external service calls, and referenced business class files. Documentation Agents helped developers generate detailed, standardized service documentation to speed up downstream engineering and reduce ramp-up time.
A Unit Testing Agent auto-generated JUnit test cases covering both positive and negative paths for targeted services like the Journal Manager Microservice, while a QE Agent began supporting QA teams with Rest Assured test script generation for API validation. The result was immediate value in the discovery phase, delivered without requiring any intrusive changes to the bank’s existing systems.
Tech Stack:
Achieved 50–75% time savings across microservices and SOA discovery tasks, proving out Agentic AI as a viable enabler for the bank's modernization roadmap.
Reduced unit test development time by 67%, freeing developers to focus on core logic instead of test case creation.
Cut documentation and formatting effort in half through AI-assisted wiki generation.
Delivered up to 6x acceleration in extracting service metadata, dependencies, and business logic.
Accelerated test preparation through automated unit and API test generation.
Positioned the bank to scale automation confidently across its broader transformation initiative in collaboration with Ascendion.
