01 The Challenge
The client, a large insurance provider, operates a complex microservices ecosystem supporting policy, claims, and underwriting workflows. As part of their quality engineering strategy, they needed to establish API test coverage across a broad surface of over 500+ APIs covering critical business transactions such as policy creation, renewals, claims processing, and payments.
However, the program faced significant constraints. Existing Swagger documentation was outdated and unreliable, making it unsuitable for automated test generation. Updated API specifications were not available, and development teams could not prioritize documentation refresh within the project timeline.
Additionally, strong interdependencies between APIs required precise execution sequencing and shared data management. A traditional manual automation approach was estimated at 120 story points making it both time-intensive and inefficient for the program’s needs.
02 Solution
HIVE team deployed a customized, AI-assisted automation approach by customizing internal AI solution HIVEQ to work directly with live system interactions instead of relying on outdated documentation. Real-time network traffic captured from UI transactions was used as the primary input, enabling accurate reconstruction of API behaviors, payloads, and response structures.
Generated test scripts were seamlessly integrated into the client’s existing automation framework, avoiding any disruption or tooling changes. To address transactional dependencies, execution flows were carefully orchestrated to ensure correct sequencing across policy, claims, and payment lifecycles.
This approach enabled rapid, reliable test generation aligned with real system behavior, while minimizing dependency on documentation and reducing manual effort.
03 Business Impact
04 Solution
HIVE team deployed a customized, AI-assisted automation approach by customizing internal AI solution HIVEQ to work directly with live system interactions instead of relying on outdated documentation. Real-time network traffic captured from UI transactions was used as the primary input, enabling accurate reconstruction of API behaviors, payloads, and response structures.
Generated test scripts were seamlessly integrated into the client’s existing automation framework, avoiding any disruption or tooling changes. To address transactional dependencies, execution flows were carefully orchestrated to ensure correct sequencing across policy, claims, and payment lifecycles.
This approach enabled rapid, reliable test generation aligned with real system behavior, while minimizing dependency on documentation and reducing manual effort.
Live traffic-driven test generation
Real HTTP request and response data captured from UI interactions used as the source of truth, eliminating dependency on stale Swagger documentation
Accurate script creation with dependency
Test scripts generated using real payloads, headers, and response structures ensuring high fidelity with actual API behavior. API sequencing and data dependencies across policy, claims, and payment workflows explicitly mapped to ensure valid end-to-end execution
Seamless framework integration
Generated scripts plugged directly into the client’s existing automation framework, avoiding rework or migration