Our client – a medium-sized insurance company in the East Coast – was experiencing major issues with administering auto policy claims, caused largely by inaccuracies in claim filing and subsequent claim settlement. One of the primary factors was the ad-hoc approach to automating the overall claims process, and the mix of desktop and web implementations was adding to the complexity


Pursuit Software performed an initial assessment which revealed the following:

  • Percentage of manual test cases automated: Approximately 35% of test cases for the web interface had been automated using standard web automation techniques, while no automation had been created for the desktop interface. Moreover, selection of automation candidates was done by focusing on certain specific areas of functionality, instead of critical, high visibility areas from an end user perspective.
  • Visibility to test progression and results: While the client was using an off-the-shelf test management tool for manual testing, there was no clear mechanism for sharing the testing statistics. This also provided a partial picture since the test suites had a mix of manual and automated tests, which were being executed using a de-facto test execution environment. Since no clear mechanism existed for providing a snapshot of the test results, test failures were not being addressed over long periods of time, thereby impacting the overall quality of the claims being administered.
Based on the findings, Pursuit implemented the following steps:

  • Created an automation road map clearly outlining the prioritized areas to be automated – covering both the desktop and web interfaces – over specific time periods.
  • Created a preliminary dashboard to display the status of manual and automated test progress, that would be accessible to all stakeholders. The dashboard incorporated specific levels of abstraction to cater to the target audience (senior management, mid-level management, engineering).


As of date, the automation level for the web interface has been extended from 35% to 65%. In addition, automation coverage for the desktop interface now stands at about 20% compared to no initial automation. In both cases, priority has been given to critical and high visibility areas determined by applying systematic test prioritization techniques to minimize the risk exposure post release.



Our client – a small hospital in the Mid-West – was experiencing major issues with its web and mobile patient portals. A large number of cases were being reported of inaccurate data in the portal related to patient lab results for multiple historical time intervals.


Pursuit Software’s initial assessment based on in-person interviews as well as independent observation of the testing process showed that:

  • Majority of the issues were resulting from a change in the development approach to use REST APIs interacting with a NoSQL backend to drive the portal functionality, compared to the robust and proven mechanism involving direct interaction with the system backend using Java and a relational database (MS-SQL).
  • A manual testing approach had been used prior to the transition to validate data insertion and response for the relative and absolute time interval combinations. However, insufficient training for the REST API calls and NoSQL database made it difficult to translate the existing test cases to the new environment, and to validate the API responses vis-à-vis the backend data.
Pursuit worked closely with the client’s testing team to:

  • Implement best practices for REST API testing, including providing support to implement a standard REST API testing tool to ensure a smooth transition to improve test execution in the new environment.
  • Implement basic test automation to help extend the testing approach to cover additional combinations of time intervals and corner cases which had not been addressed previously.


Some of the benefits post implementation were:

  • Test coverage was extended by 22%.
  • Test execution time was reduced by 15%.
  • Number of defects reported was reduced by 35%.