Leapwork hands off code validation to AI agents to keep pace with automated software development

Danish agentic artificial intelligence startup Leapwork ApS has announced the launch of its fully automated Continuous Validation Platform, aiming to help enterprises keep up with the velocity of generative AI software development.

The rise of AI “vibe coding” means that software development has accelerated to breakneck speed, enabling enterprises to deploy new code faster than ever before. That’s great for innovation, but it dramatically increases the risk of teams accidentally breaking something.

Leapwork’s new Continuous Validation Platform is meant to prevent that. To keep pace with the volume of AI-generated code being shipped, it too is a fully automated “agentic” platform that aims to provide a kind of safety net for developer teams. The idea is to ensure every bit of code, whether written by humans or AI agents, is validated at every step of the development lifecycle and in every environment.

The startup is taking aim at a widening “trust gap” in software engineering. Citing its own research, it says 88% of software engineers now prioritize AI. That’s having a dramatic and negative impact on their quality assurance budgets, which have swollen from 35% to 40% of their total spend, on average.

The issue is that traditional code testing tools cannot keep pace with the compressed release cycles of AI-generated software. Bots can generate new code in minutes, compared with hours for human developers, yet their outputs can be a lot more unpredictable. So when AI builds software, there needs to be more emphasis on validation – proving that the code works in a deterministic and auditable way so it can be safely deployed in production.

Chief Executive Kenneth Ziegler said the Continuous Validation Platform is “application-agnostic,” which means it can be used to validate code across modern web applications, mobile and even legacy systems such as Oracle and SAP. The platform consists of three main components, including Leapwork Studio, which is a no-code engine featuring “AI Automation Blocks” that aim to validate if AI outputs are correct and visualize data in the same way as humans do.

The second component is Leapfrog Performance, which integrates performance and application programming interface testing into the software development workflow. With this, engineering teams can import their existing JMeter scripts into their workflows and use them to model real-world traffic scenarios.

Finally, Leapwork AI Studio is intended to be the “agentic” heart of the platform, built to validate code written by tools such as Claude Code and Cursor. It’s based on an intelligent knowledge base that ingests newly generated code and designs the test plans and scripts needed to validate it.

Ziegler said Leapwork wants to become the “deterministic foundation” of the automated software development era. He said developer teams are excited about agentic automation, but are also facing “unparalleled anxiety” over the stability of their AI-generated code.

“Continuous validation is expensive and tricky to figure out without domain expertise, history and experience,” he said. “With the Leapwork platform, our customers get access to the velocity and intelligence of AI-native automation alongside proven deterministic execution that enterprises need to meet their quality and governance requirements.”

Image: SiliconANGLE/Meta AI