News

AI Enhances Tool for Automated Software Test Creation

• By John K. Waters
• 10/16/2018

The advent of artificial intelligence (AI) and machine learning is changing the software development process, sometimes in radical ways. When it comes to software test automation, the focus so far has been on enhanced test execution, while the test creation process has remained largely manual. The status quo changed last week with the latest release of Parasoft SOAtest, which comes with new machine learning techniques that augment the tool's Smart API Test Generator.

SOAPtest is Parasoft's API and UI functional testing solution. The AI-and-machine-learning-powered testing tool is designed to transforms existing test artifacts into security and performance tests, the company says, and build a foundation of automated tests that can be executed as part of Continuous Integration and DevOps pipelines.

Introduced in May, the Smart API Test Generator took a big step in this direction by providing organizations with the ability to create meaningful API test scenarios automatically using AI to understand how their APIs are being used as their applications are exercised from the UI.

The new feature allows users to leverage their organizations' existing test inventories and train SOAtest's AI engine using ML, so it can create meaningful API test templates automatically. This new feature is designed to enable organizations to build a foundation of tests quickly to increase API coverage and scale their API testing practice.

The company showcased the new release recently at the Starwest software testing conference in Anaheim, Calif., during a talk with industry analyst Theresa Lanowitz. The founder and head analyst at market research firm voke, Lanowitz is one of the leading strategic thinkers on automated testing.

The new SOAtest capability fits Lanowitz's recommendation that organizations focus on reducing all manual testing efforts to support Continuous Quality.

"Over the past decade or so we have seen testers attempt to create more automated UI tests," Lanowitz told PureAI.com. "The problem with these UI tests is that they are brittle, expensive to maintain, and time-intensive to create. There's also quite a bit of manual testing at the UI level. In either case, automated or manual, testers are unable to keep pace with development and the rework/maintenance costs of tests is quite high."

By allowing AI and ML to work with an existing catalog of tests, she explained, the Parasoft product allows testers to become immediately more efficient and test at the API layer to detect defects earlier, achieve higher quality, and reduce maintenance and rework of tests.

"This in the type of innovation we need in the testing market to empower testers to be strategic players in the software engineering lifecycle and help protect the brand and the business," she said.

The new version of SOAtest bolsters AI-driven test creation by allowing testers to train the Smart API Test Generator to create test scenarios based on the rules an organization uses to define how it tests its business logic. By adding this additional logic into the traffic-parsing process, the tool learns the relationships among different test steps, so it can intelligently connect them into scenario-based regression suites. "Leveraging this technology ensures that API test inventories will be efficient, consistent, and change-resistant," the company said.

More information is available on the Parasoft Web site here.