Tricentis unveils agentic AI Workspace for safer releases

Tricentis has launched an "agentic" software quality engineering platform that brings AI-driven testing and quality controls into a single workspace, with governance and human oversight built into day-to-day execution.

It is positioning the product around a growing concern in large organisations: AI-assisted software development can increase the pace of change faster than established testing and release processes can absorb. Tricentis argues this imbalance raises the risk of defects reaching production systems and triggering outages across interconnected applications.

That risk has measurable business impact. In 2025, enterprises faced a median cost of about GBP £1.5 million per hour for high-impact IT outages, according to figures cited by Tricentis. In the UK, the Treasury Committee has also warned that current approaches to AI in financial services could cause "serious harm to consumers and the wider system", and has called for safeguards.

AI workspace

At the centre of the release is Tricentis AI Workspace, described as a unified command centre for quality engineering. It is designed to coordinate multiple AI agents across testing, automation, performance and quality intelligence, with governance features such as approvals and auditability embedded into execution.

The platform targets large enterprises with mixed technology estates, drawing on Tricentis technology across nearly 200 enterprise resource planning systems and packaged applications. It also supports web and custom applications.

Tricentis is pitching the platform as an alternative to generic AI tooling for software engineering. It argues general-purpose tools can lack application-specific context and an understanding of dependencies between systems, making outputs unreliable in complex enterprise environments.

Team of agents

The platform includes several AI agents, each with a defined role across the software development lifecycle. Tricentis Agentic Quality Intelligence is designed to interpret change, risk and quality signals and determine release readiness. It can direct testing and escalate issues to humans when judgement is required, according to the company.

Tricentis Agentic Test Automation updates an earlier product. Additions include support for SAP GUI and web applications, deeper integration with Tricentis Tosca automation engines, and reuse of test modules to reduce duplication and maintenance work.

Tricentis Agentic Performance Testing focuses on performance validation, embedding autonomous agents across analysis, design and execution. Tricentis claims this can accelerate insights by up to 90% to 95% and reduce reliance on specialist manual work.

Tricentis Agentic Test Creation integrates with Tricentis qTest. It sits alongside test engineers for authoring and includes natural-language test creation. Tricentis says it can generate reusable test cases faster and more consistently, while reducing duplication and reliance on specialist expertise.

Governance focus

Control and accountability are central to the pitch. Policy enforcement, approvals and auditability sit in one place, with human review gates and oversight built into execution. Tricentis frames this as a way to scale AI use without increasing compliance or operational risk.

Another theme is accessibility for quality teams. Tricentis says the platform enables "zero-code" creation and management of AI agents, reducing the need for specialised skills or custom development when rolling out AI-driven quality processes.

"AI is transformative in its ability to create code at unprecedented speed, however the friction caused by lack of confidence in the quality of the output is causing CIOs real pain. While enterprises demand speed, they also can't afford to introduce risk through unsecure or low-quality AI-generated code," said Kevin Thompson, Chief Executive Officer, Tricentis.

"That's the problem Tricentis is solving today. We're offering the first end-to-end agentic software quality platform that redefines how enterprise software can be tested, governed, and released to deliver high-quality code at the speed of AI while safely accelerating time-to-value," Thompson said.

Tricentis is also pointing to its own internal deployment as evidence, arguing agentic testing can compress delivery timelines in major programmes such as migrations.

"We're already using agentic testing at Tricentis and are experiencing real impact in our transformation projects," said David Cowell, VP of AI and Machine Learning, Tricentis.

"A cloud migration that would typically take a few months took us just one week with agentic AI. That's the kind of step-change enterprises need, compressing release cycles without increasing risk, and enabling teams to move faster without cutting corners on quality," Cowell said.

Customer input

Wolters Kluwer has beta-tested the platform, according to Tricentis. The publisher and information services provider described the release as part of a broader shift in how quality work is organised as AI increases software change volumes.

"The Tricentis AI Workspace is a strategic inflection point for us because it enables agentic orchestration across the entire software development lifecycle - not just code generation or test automation," said Paul DiGrazia, Vice President of Quality Engineering, Wolters Kluwer.

"As AI accelerates software creation, the real challenge becomes trust - software is often nearly right, but not production-ready. Our approach shifts Quality from manual test creation to confidence engineering. Instead of simply generating tests, our agents actively identify unknowns, prevent defect classes, orchestrate risk-based validation, and provide readiness signals that support human decision-making. This allows us to move at the speed of AI while shipping responsibly - reducing risk, improving release confidence, and delivering measurable ROI through faster delivery, fewer escapes, and continuous learning at scale," DiGrazia said.

Tricentis says the agents in AI Workspace form the basis of a broader platform that will evolve over the next several years towards more autonomous quality engineering with continuous governance.