Testkube Launches AI Test Creation, Closing the Gap Between AI-Written Code and Tested Software
Testing platform provider Testkube has launched AI Test Creation to enable teams to describe a test in plain language and receive a working test in the framework they already use, and run it in their own infrastructure before they accept it.
Developers now write and ship more code than ever, much of it drafted with AI, but they still create tests the way they always have: someone writes the test, wires it into a repository, adds it to a pipeline and finds a real environment to run it against. That work rarely reaches the top of a sprint, so coverage falls further behind the code with every release. Using AI tools like Claude or CodeX only solves half the problem. Instead of being limited to isolated and local test execution, AI Test Creation creates and runs tests inside your infrastructure, then opens a pull request in users’ GitHub repository so they can review, edit and version the test like any other code. Teams that need execution data to stay inside their own cluster can run all of it on-premises.
“I’ve spent 20 years watching teams struggle to keep their testing in line with committed code, and AI is widening that gap even further,” said Ole Lensmar, co-founder and chief technology officer of Testkube, said in the announcement. “It’s not just about creating tests, it’s everything that happens after to make those tests work: the wiring, the environment, the results.. This is what we’re building: AI that uses your existing tools and creates tests that are immediately integrated into your pipelines and proven in your real infrastructure. “
AI Test Creation includes:
-
Any framework, any test type. Tests are generated in the frameworks a team already uses, across end-to-end, API, load,infrastructure testing and more, with skills built for the most widely used tools and scenarios.
-
Immediate execution. Every generated test runs in the team’s real environment within seconds, so a wrong assumption surfaces while it’s still a draft.
-
Tests the team owns. Accepted tests arrive as pull requests in the team’s GitHub repository, reviewed and versioned like any other code.
-
On-premises deployment with your models. Tests and execution data stay inside the customer’s own cluster, using the LLMs you provide.
Get started here.
The post Testkube Launches AI Test Creation, Closing the Gap Between AI-Written Code and Tested Software appeared first on SD Times.
Tech Developers
No comments