Take a faster, smarter path to AI-driven C/C++ test automation. Discover how >>
WEBINAR
Software delivery today can feel like a puzzle. Third-party systems, legacy infrastructure, and complex integrations create delays, bottlenecks, and unpredictable test environments. But what if you could retake control and speed up delivery without sacrificing quality?
Velera and ING Mortgages did. Velera doubled unit test generation and achieved 85% code coverage, and ING Mortgages made 60% of their complex mortgage ecosystem testable while cutting test preparation from five days to just one. All in complete isolation, with data dependencies and environments fully under their control.
What did they have in common? Service virtualization. And in this session, industry experts from both companies are here to talk about how they went about implementing it.
Learn about how you can use service virtualization to:
Every development team runs into testing roadblocks. It could be waiting for third-party systems, dealing with limited data, or managing complicated connections. When you can’t fully control your testing environment, things just slow down.
Hirakant from Velera, a fintech company serving credit unions and community banks, shared how their software delivery heavily relies on other fintech companies and data providers. Their APIs need to be available in lower environments for testing. However, non-production environments don’t always have the same reliability as production. This directly impacts delivery dates and project costs.
Hirakant recalled instances where delays stretched over a month, causing a ripple effect upstream for teams dependent on Velera’s APIs. The impact on development time, testing, cost, and overall quality was significant.
He shared a few specific pain points:
These issues pushed Velera to seek a better way of working.
Marcel from ING Mortgages in the Netherlands described similar challenges. They operate in an agile environment where API owners constantly change versions. Aligning with all teams to get the acceptance environment in a specific setup for testing was a major effort.
ING Mortgages also uses third-party software for their mortgage application, treating it like a black box. This black box needs to connect to all its interfaces. To make things more complex, it’s very state-dependent, meaning they can’t test individual interfaces in isolation. They have to test the entire system to see how the workflow handles all interactions.
Creating virtual services that accommodate their testing took significant effort. However, by owning the data, they could ensure all test cases and scenarios were covered. This led to much higher code coverage for their regression tests and made their work much easier, increasing confidence in their software releases.
Service virtualization was a new concept at Velera. Hirakant pushed for it, driven by the pain points mentioned earlier. His search for a solution led him to Parasoft after evaluating other providers like SmartBear, IBM, and Traffic Parrot.
What made him choose Parasoft included:
Marcel joined ING after service virtualization was already in place, and Parasoft had been used there for some time. Around 2020, the mortgages department also adopted Parasoft. Marcel, having worked with similar virtualization products before, found Parasoft intuitive and comfortable to use. He noted its rich toolset, which covered all their needs and was easy for colleagues to understand and modify.
Parasoft’s virtual services encapsulate functionality, making it easier for business analysts to understand the data and how the functionality works. This leads to better predictions for test outcomes. For example, when developing APIs for a new release, they can conduct an experimental phase by introducing new elements and observing how the mortgage application responds. This speeds up the development cycle.
One of the most noticeable impacts for Velera was achieving code coverage exceeding the 80% limit, which was a major blocker before. Now, creating virtual assets is part of their process when dealing with third-party integrations. This allows for parallel development activities, compressing timeframes significantly.
Previously, finalizing an interface design, waiting for third parties to make their environments available, and then starting testing could take a long time. With service virtualization, they could work in lower environments using virtual assets even before the third party’s environments were ready. This led to a high degree of confidence that things would work as agreed upon.
For projects that might have taken a year to complete end-to-end, service virtualization helped them achieve this in just 3 to 4 months – a massive improvement.
ING’s goal is to work in an agile manner, and they increased their release cycle from one month to two weeks. This requires the right tools. A combination of an Azure release pipeline, a robust regression test set tightly coupled with Parasoft virtualization, has sped up their release cycle and significantly improved the quality of releases.
Business production interruptions have gone down drastically. Before, they had multiple minor issues per year, but now they have significantly fewer, leading to better control. Manual testing has been replaced with solid regression cases combined with virtual services, allowing them to rely more on code quality and detect issues early.
Hirkant added that with the move to stricter agile modes and self-contained teams, virtual assets eliminate the need to wait for other teams to manage environments or test data. Teams can create or reuse virtual assets as needed. For API platform upgrades, which previously required significant effort to set up test environments and repeat testing, coupling service virtualization with their test automation framework makes regression testing as simple as pushing a button, reducing tasks that took days or months to minutes or hours.
AI is bringing a lot of excitement to software development and testing. Hirkant is excited about Parasoft’s new AI capabilities for generating virtual services. While Velera hasn’t adopted it yet, they see a big need for AI in creating test cases by feeding in requirements and outcomes.
Marcel attended a Parasoft customer event where AI and LLM possibilities in virtualization were demonstrated. He believes AI can make work easier and faster. His main concern is verifying the correctness of AI-generated outputs. ING is exploring AI in various tracks to improve business processes, including their DevOps environment.
Both speakers agree that AI has the potential to make testing more autonomous and efficient, from test creation to generating working virtual services, making it easier to adopt and scale technologies across teams.
