Summary
At Tricentis Transform on August 20, 2026, Tricentis introduced three AI capabilities for quality engineering: Aida, an autonomous agent that explores web and Windows desktop apps to surface defects and coverage gaps without any pre-existing test suite; AgentScore, which shifts from deterministic testing to probabilistic evaluation of how AI agents behave in real workflows, with review/block/ship recommendations; and Release Risk Intelligence, which highlights release-level coverage gaps and suggests next actions.
What changed
Tricentis announced Aida (script-free autonomous exploration of web and Windows desktop apps to find defects and coverage gaps), AgentScore (probabilistic, workflow-behavior evaluation of AI agents producing composite quality scores and review/block/ship guidance), and Release Risk Intelligence (release-level risk and coverage-gap analysis with recommended actions), developed through the Tricentis Labs incubator.
Why it matters
As teams ship software built and run by AI agents, testing has to move from deterministic pass/fail scripts toward probabilistic evaluation of agent behavior; AgentScore is a direct attempt to make that shift operational. It positions quality engineering as a control point for trusting agent output before release, a growing enterprise need as agentic coding spreads.
Evidence excerpt
Tricentis Aida: an autonomous agent that explores web and Windows desktop applications to surface defects and coverage gaps without any pre-existing test suite or scripts. AgentScore: shifts quality engineering from deterministic testing to probabilistic evaluation, generating composite quality scores with review, block, or ship recommendations. Release Risk Intelligence: highlights release-level coverage gaps and suggests actions.