Tricentis has introduced three AI technologies through Tricentis Labs, an innovation program focused on testing emerging technologies with customers and partners before turning the most promising concepts into enterprise products.

The technologies—Tricentis Aida, Tricentis AgentScore and Tricentis Release Risk Intelligence—target different stages of the software development lifecycle. Together, they reflect a broader shift in quality engineering from scripted testing toward systems that can explore applications, evaluate probabilistic AI behavior and help engineering teams decide whether software is ready for release.

The announcements were made at Tricentis Transform, the company's annual industry conference. They arrive as enterprises increasingly use generative AI and AI coding tools to accelerate development, creating pressure on testing organizations to keep pace with a much faster software delivery cycle.

That creates a fundamental problem for traditional quality processes. Conventional automated tests are typically designed around expected application behavior. AI-powered systems, by contrast, can produce different outputs depending on context, prompts, data and interactions. An enterprise therefore needs to evaluate not only whether a system passes predefined tests, but also how reliably it behaves across real-world scenarios.

Tricentis Aida is designed to address the application-exploration side of that challenge. The AI agent can autonomously explore web and Windows desktop applications, identify potential defects and highlight coverage gaps without requiring an existing test suite, scripts or extensive setup.

The concept is important because test creation itself can become a bottleneck. If an AI agent can navigate an application and identify areas that deserve testing, quality teams could spend more time reviewing risk and less time manually constructing initial test coverage.

Aida also represents a move toward autonomous testing rather than simply AI-assisted testing. Instead of helping an engineer write a test case, the technology is intended to explore the application independently and return information about application health.

AgentScore tackles a different problem: how to determine whether an AI agent is actually ready for production.

Tricentis describes AgentScore as a shift from deterministic software testing toward probabilistic evaluation. The technology observes AI agents operating in real-world workflows, recommends what should be measured and generates composite quality scores. It can then provide recommendations to review, block or ship an agent.

That approach reflects one of the industry's biggest unresolved challenges. A conventional application can often be evaluated against known expected outputs. An AI agent may achieve the same business objective through different paths, generate different responses or make decisions based on changing context.

For enterprises, that makes AI quality less binary. Accuracy, consistency, safety, task completion, policy compliance and behavior under unexpected conditions may all need to be evaluated.

AgentScore's proposed composite scoring model is therefore aimed at turning those variables into a decision framework that engineering and business stakeholders can understand.

The third technology, Release Risk Intelligence, focuses on the final stage of the software delivery process. Rather than asking simply whether a release passed its tests, the capability is designed to identify what remains exposed to risk.

Tricentis says the technology surfaces release-specific coverage gaps, prioritizes risks by severity and recommends actions through contextual AI assistance. For release managers and quality leaders, that could provide a more focused way to assess whether unresolved issues are significant enough to delay a deployment.

The three technologies collectively point toward a different role for quality engineering. Instead of operating primarily as a verification stage at the end of development, quality intelligence is increasingly being embedded throughout the lifecycle.

That evolution is happening alongside rapid growth in AI-assisted software development. Gartner has predicted that by 2028, 75% of enterprise software engineers will use AI code assistants, up from less than 10% at the beginning of 2023. The research firm has also warned that organizations need stronger governance and validation as AI becomes embedded in software engineering workflows.

That forecast helps explain the strategic direction behind Tricentis' investment. If software development becomes substantially faster through AI coding agents, testing cannot remain dependent on processes designed for slower, human-led development.

The company's acquisition of Tabnine adds another piece to that strategy. Tabnine provides AI-powered software development capabilities, including enterprise-focused coding assistance. Tricentis announced its agreement to acquire Tabnine in 2025, describing the deal as a way to strengthen its broader agentic quality engineering strategy.

The combination could eventually give Tricentis a broader position across the AI-assisted development lifecycle: helping developers create software while using AI-driven quality engineering to explore, evaluate and release it.

That puts Tricentis in a competitive market that includes established software testing providers such as SmartBear, OpenText and IBM, as well as newer AI-native development and testing platforms. Microsoft, GitHub and other major technology companies are also pushing AI coding agents deeper into enterprise software development.

Tricentis' differentiation is increasingly centered on quality as an intelligence layer across the development lifecycle. Its argument is not simply that AI can make testing faster, but that AI needs to become part of the quality decision itself.

For enterprise engineering organizations, that distinction matters. The risks associated with AI-generated software are not limited to conventional defects. Organizations must also consider hallucinations, unpredictable agent behavior, security vulnerabilities, data exposure and failures that emerge only under unusual conditions.

A quality platform that can continuously evaluate those risks could become a critical control point as enterprises move toward agentic software development.

The challenge will be proving that these AI systems themselves can be trusted. Enterprises will need transparent scoring methodologies, explainable recommendations, auditability and human oversight before allowing AI-generated quality decisions to influence production releases.

Tricentis Labs is effectively using its customer base as an early testing environment for that future. Its model allows emerging capabilities to be exposed to real-world enterprise requirements before broader productization.

The bigger story is that software quality is becoming a moving target. As AI agents increasingly write, test and operate applications, enterprises will need quality engineering systems capable of evaluating not only software but the behavior of the agents building and interacting with it.

Market Landscape

Software testing is undergoing a transition from deterministic automation toward AI-assisted and increasingly autonomous quality engineering.

Traditional testing platforms from vendors such as Tricentis, OpenText, IBM and SmartBear remain focused on automation, test management and application quality. At the same time, AI development platforms from Microsoft, GitHub and other technology companies are changing how applications are built.

The emerging competitive category sits between these worlds. AI agents can generate code and perform development tasks, but enterprises need independent mechanisms to determine whether the resulting software is reliable enough for production.

Tricentis' Aida, AgentScore and Release Risk Intelligence address three different pieces of that problem: discovering application risks, evaluating AI-agent behavior and assessing release readiness.

The opportunity is substantial, but so is the validation burden. Enterprises are unlikely to accept opaque AI quality scores for mission-critical software without evidence showing how those scores are calculated and how reliably they predict production risk.

Strategic Outlook

The quality engineering market is likely to become increasingly important as agentic software development moves from experimentation into enterprise production.

AI coding assistants can accelerate development, but faster code generation creates a corresponding need for faster validation. Gartner's forecast that 75% of enterprise software engineers will use AI code assistants by 2028 illustrates the scale of that transition. (gartner.com)

The next generation of quality platforms may therefore function less like traditional testing suites and more like continuous intelligence systems. They will need to understand application context, business processes, release changes and AI-agent behavior simultaneously.

Tricentis' strategy—including its Tabnine acquisition and Tricentis Labs program—suggests the company wants to occupy that broader control layer.

For enterprises, the winning technology will ultimately be the one that can increase development velocity without sacrificing confidence. AI can make software creation faster, but the ability to prove that software is safe, reliable and ready for production will determine how far agentic development can scale.

Top Insights

• Tricentis Aida autonomously explores web and desktop applications, helping quality teams identify defects and coverage gaps without traditional test scripting.
• AgentScore introduces probabilistic AI-agent evaluation, giving enterprises a framework for measuring agent behavior and deciding whether systems are production-ready.
• Release Risk Intelligence uses AI to identify coverage gaps and prioritize release risks, helping engineering leaders make faster deployment decisions.
• Gartner expects 75% of enterprise software engineers to use AI code assistants by 2028, increasing demand for automated quality and governance.
• Tricentis' Tabnine acquisition expands its strategy from software testing toward an integrated agentic quality engineering platform spanning development and release.

Get in touch with our MarTech Experts