Cyara Launches Agentic Testing

Cyara, a provider of customer experience (CX) assurance solutions, has added agentic testing and artificial intelligence governance capabilities to validate, monitor, and control AI agents across voice and digital channels before and after those agents interact with customers.

"Every enterprise wants to deploy AI agents in their contact center. The ones who actually will are the ones who can prove those agents work, before customers find out they don't," said Sushil Kumar, CEO of Cyara, in a statement. "We built these capabilities because the level of assurance has to match the level of autonomy. If you're putting an AI agent on a live customer call, you need to know it will handle the conversation correctly, comply with regulations, and not introduce bias. That's what Cyara now delivers."

Cyara's latest capabilities are purpose built for agentic AI in voice and interactive voice response (IVR) and include the following:

• Agentic AI Testing for Voice & IVR, which provides an assurance layer that tests AI agents with AI agents to catch what scripts can't and detects failures and regressions before deployment and into production.
• AI Trust, Cyara's testing suite for optimizing generative AI-powered CX and the governing layer for AI CX, has expanded through two new modules, Compliance and Bias. The modules identify compliance risks and detect bias that could impact customer outcomes and brand trust. Cyara's AI Trust suite provides governance capabilities for generative AI systems, including validation for accuracy, bias, misuse, and regulatory compliance. The new Compliance and Bias modules expand this coverage to help organizations detect ethical and legal risks before they reach customers.
• A recommendation engine for agentic CX, which assists teams in designing and optimizing prompts and accelerates test development. It helps users generate prompting strategies for agentic and hybrid tests, combine scripted and agentic approaches when designing test cases, and improve test coverage and reliability.