Curiosity Software Unveils Quality Modeller®

• Oxford, UK – August 2026 -

• Oxford, UK - August 2026. Curiosity Software today announced Quality Modeller®, a platform built to address the longstanding, unsolved problems of poor and slow test case design and to provide a measurable framework to store, build, and manage your test cases.

  Test case design has gone through several evolution from manually created to constrained combinatorial an onto controlled natural languages, including Gherkin. Model-based testing marks the next major shift, and Quality Modeller® represents the latest evolution.

  By creating a decomposed and structured BPMN model of a requirement along with business constraints and expected results, the combination of AI and powerful mathematical path analyser can create unique combinations of test cases. These test cases can be expanded and adjusted to suit the type of tests required, from Manual to automated.
  
  Critically the ability to update the requirements and automatically repair, discard and create new test cases save an enormous amount of time.
  
  

  The Problems That Never Went Away

  Despite decades of tooling change, the core problems with test cases have persisted. Test cases accumulate indefinitely rather than retiring with the logic they once covered. Manually authored cases are often duplicates rather than genuinely new coverage. Teams typically have no reliable way to measure what percentage of real application logic is being tested. Furthermore, a small requirements change can force every test case to be checked whether it is relevant or not. Realistically manual test case design can produce only around six or seven test cases a day, a hard ceiling that doesn't scale with modern release cycles.

•  

•  

  Why AI Alone Doesn't Fix It

  As AI capabilities have grown, many teams have looked to AI to close these gaps. Curiosity Software's position is more measured: AI is genuinely helpful, but it also gets things wrong, and it cannot be relied on alone to build test cases without oversight.

  For AI to be trusted in this process, it needs to be observable. AI created test cases need to be checked thoroughly. This is where visual modelling excels, human brains process visual information more naturally than text and crucially multiple team members can quickly confirm that a model, and the constraints layered onto it, genuinely match with the requirements. Model Based Visual design create the common understanding that is crucial to designing the systems the users want.
  
  Once the model is agreed, AI accelerators can help validate and enhance the models checking for boundary cases and any compliance standards have been met. Once complete the test case can automatically be generated.

•  

•  

  One Model, Every Stakeholder

  Quality Modeller® is designed to give business analysts, developers and testers a single visual, searchable source of truth, replacing the separate documents, spreadsheets code, scripts, and schemas each function traditionally uses. Teams can use the models to visually define complex processes, breaking them down into common components, surface gaps in understanding, and align every stakeholder around one shared source of truth. Once complete you can generate the assets, including: tests, data, and automation early in the development cycle.

•  

•  

  AI Accelerated, Human-in-the-Loop

  Quality Modeller®AI is applied on top of validated models rather than in place of them. The platform uses AI to instantly generate covered tests from requirements, match test data to those requirements, and run quality reviews that flag gaps and areas of low coverage. Because the underlying models are structured and visual, Curiosity Software also positions them as a foundation for training AI systems in ways that reduce knowledge silos across an organization.

  In one deployment, a global accounting firm reduced the time required to convert a requirement into test cases from roughly 7 hours to just 45 minutes using Quality Modeller®'s AI accelerated workflow, while keeping humans responsible for confirming that the underlying model was correct.

   

•  

• Build your Models from existing Assets

  Implementing Quality Modeller® does not mean you need to start from scratch. Simply load in any existing requirements, Test Cases, Gherkin scripts, code into Quality Modeller. You can then build you visual models and the tool will give you an accurate measure of how good current tests are.

   

"AI can be very helpful, but you can't rely on it alone," said Harry Burn, Head of Solution Engineering at Curiosity Software. "To trust what AI produces, it has to be observable. That's what a visual model gives you a way for humans to quickly confirm the model and that its constraints match the requirements. All of this before that model is used to generate a single test case."

Quality Modeller® is available now as part of the Curiosity platform. Organizations looking to apply the approach to their own test design process can contact Curiosity Software to learn more.

Media Contact

Sophia Ryan

Marketing Executive

Curiosity Software

Sophia.Ryan@curiosity.software

www.curiositysoftware.ie