Wolters Kluwer drives consistent software quality at scale with agentic AI

Wolters Kluwer started almost two centuries ago and was originally a print company. Over the past few decades, the organization has been on a sweeping digital transformation journey and introduced new technologies that rolled out at a rapid pace. The company has grown exponentially over the last few years, and a large part of that success can be attributed to ensuring that everything it delivers to customers is of high quality.

With the influx of AI in the org’s toolset, particularly applied to software development, Paul DiGrazia, Vice President of Quality Engineering at Wolters Kluwer, set out to build a quality engineering strategy to match the increasing volume and velocity of AI-generated code – and to catch the errors that AI can sometimes introduce – and that are difficult to catch with traditional testing methods. To that end, he turned to qTest’s agentic AI capabilities to help his team create test cases faster and ensure broad coverage across diverse use cases.

DiGrazia and his team of veteran qTest users were early adopters of the AI, quickly deploying to 400 users within the Wolters Kluwer quality engineering org. In a short period of time, the org achieved 30% time savings in test case design by directing agentic AI to draft test cases, steps, and expected results from user requirements. As Wolters Kluwer moves toward a fully integrated development environment (IDE), where debugging takes place right alongside code creation, the ability to accelerate quality with AI is pivotal to success.

DiGrazia firmly believes quality metrics are vital to maintaining business continuity. “From a quality engineering perspective, the most effective metric is our escaped defect. We also look at our customer experience metrics and have a preventative action process aligned with it,” said DiGrazia.

DiGrazia and his team oversee quality standards of processes and tools over a wide scope of applications and platforms and the technology behind them.

“The quality engineering org is responsible for hundreds of applications and probably half a dozen major platforms. We did a big cloud migration, so it’s everything from Azure and AWS and web technologies like Java, .Net, databases, SQL, and Oracle. We have newer platforms and products and some older technologies that have been around for 20 years,” shared DiGrazia.

qTest has been critical for DiGrazia’s team’s sprawling application test management, and DiGrazia looks to incorporate Tricentis AI Workspace into the test management strategy as AI usage increases at all stages of the SDLC. He expects AI Workspace “will make a big impact as we move to evolving quality an entirely autonomous process at Wolters Kluwer.”

In an AI Workspace trial, DiGrazia created a coordinated escaped defects tracking workflow across multiple AI agents. “Within an hour, we had a whole agentic escaped defects tracking process outlined,” DiGrazia says, which is automatically documented in qTest.

Having this level of visibility is a major plus for DiGrazia and team, since it allows for close tracking of the important metrics they depend on for go/no-go release decisions.

DiGrazia’s test automation philosophy is based on the typical test automation pyramid. “Years ago, Tricentis came up with a nice charter about the pyramid that I use all the time because it shows where you want to make the heavy investment on your API level and unit testing and then make sure that you can get the right amount of UI testing that’s going to provide that return on investment. That’s how we’ve been able to operate fast for all of our test automation.”

Wolters Kluwer uses an open-source framework that integrates with an in-house tool and BlazeMeter for performance. For testing, the organization used everything from manual spreadsheets to Word documents to tools such as Jama. It needed a test management solution to keep all the testing sources unified and centralized and to start adding more test automation.

Wolters Kluwer selected qTest as its test management solution to maintain visibility and control across a diverse test automation toolset. The team uses qTest to orchestrate automated testing and to standardize, centralize, and scale testing best practices across the enterprise.

With this unified approach, the team has achieved 100% automation at the UI and API levels for its core platform applications. For many others, their automation rates range from 30% to 60%, with a focus on repeatability to maximize their ROI.

“We focus automation on some of those highly matrixed and repeatable types of regression tests. That ensures we can free humans up for finding the negative test cases, and then we can get through our test cycles more quickly without having this massive regression burden,” said DiGrazia.

“qTest is where we drive all our testing metrics from in terms of test execution, test development, and percent of automated tests. We leverage the qTest APIs to feed all that data into an internally home-grown tool married with Jira. We have a full picture of tests run and what the quality is on the other side,” said DiGrazia.

Wolters Kluwer has invested in training to instill consistency across its divisions, but DiGrazia said, “There is no way that we could scale if we didn’t have some of the tools like qTest.”

After incorporating qTest enterprise AI into their quality toolchain, DiGrazia hosts regular agentic AI training sessions and is developing a set of guidelines and standards for training AI models and using AI tooling, including optimized prompting techniques to improve AI outputs. He and his team have collaborated closely with Tricentis to maximize the effectiveness of this training and develop materials to scale AI use across the entire 400-person user base.