Intelligent UI Test Automation
Let Exploratory Bots Test Your Website
Automated bots scan web UIs, generating rigorous test cases, scripts and data. Test Modeller combines the speed and ease of AI-driven test creation with the granularity and full control of model-based test automation. Exploratory bots eliminate test scripting bottlenecks, building tests and visual models automatically from web UIs. The intuitive flowcharts then maximise visibility and provide a central point of collaboration, allowing cross-function teams to focus testing where it matters most.
A new approach is needed in UI test automation
Organisations today rely on their web front for gaining and retaining business, with customers who will quickly look elsewhere when a seamless user experience is not delivered. Test teams are therefore under pressure to ensure that fast-changing web UIs remain high-performance and bug free. They depend on test automation to execute a high volume of tests in in-sprint. However, UI test automation typically relies heavily on test scripting, creating brittle tests that break after UI changes. Testers must then analyse complex applications to update existing tests after each change, before data provisioning bottlenecks create further delays. UI testing in turn falls behind releases, leaving business-critical UIs exposed to bugs. A new approach is needed to UI test automation.
Bot-driven testing generates UI tests where they matter most
Test Modeller enables in-sprint UI testing, providing exploratory bots that generate rigorous automated tests and data. Testers can target a website with a single click or Slack message, triggering high-speed bots that scan pages for elements, identifiers and logic to test. The bots further identify test data needed to test the webpage. If a new field is encountered, test teams can quickly specify the positive and negative data required for rigorous testing, with the bots learning which data to use in future. Within minutes of triggering the bots, testers can run the generated tests in their existing frameworks and CI/CD pipelines. Meanwhile, Test Modeller has updated test cases in test management tools.
The intelligent bots provide everything needed for in-sprint testing, without losing time to DOM analysis, test data delays, or test script maintenance. The bots also build visual models of the website under test, introducing all the control and rigour of model-based testing. Testers and business users can collaborate closely from the visual flowcharts, picking high-risk areas of the system to home in on during testing. This triggers exploratory bots to target the recently changed or previously buggy functionality, auto-generating optimised tests for in-sprint test automation. With Test Modellers, test teams enjoy all the speed and ease of AI-driven test creation, while retaining the visibility and granularity provided by model-based test prioritisation.
Robotic Test Design: The speed and ease of AI, the precision of models
This short demo showcases bot-driven test generation for an ECommerce store, search engine and UI with Shadow DOM elements. You will discover how:
Exploratory bots from Test Modeller rapidly scan web UIs, automatically generating test automation scripts for existing homegrown, commercial and open source frameworks.
The bot-driven test design identifies the locators and identifiers needed to generate accurate page objects, as well as the UI logic required for in-sprint test automation.
The robotic test generation identifies the test data needed to test each element in-sprint, learning from rapid tester input when an unrecognised element is detected.
The high-speed exploratory bots also build intuitive visual flowcharts of the website under test, providing a collaboration point for testers, developers and business analysts.
The visual flowcharts introduce full control to AI-driven testing, allowing cross-functional teams to fine tune testing and visualise exactly what has been tested in previous runs.
Measurable test optimisation focuses test generation on new or high-risk areas of the system, picking system logic in visual models to trigger high-speed exploratory bots.
The “zero touch” test generation and execution pushes rigorous tests to existing test automation frameworks and test management tools, while triggering CI/CD processes.
Previously unfeasible test maintenance becomes as quick and simple as reselecting a page to test, while self-healing page objects minimise delays associated with brittle test scripts