What changed
NiCE Ltd. launched its next-generation Cognigy platform, emphasizing operational reliability over raw AI intelligence. The release includes a redesigned Agentic Experience Studio, AI Assist for agent development, reusable skills, turn-by-turn tracing, and native voice-to-voice model support. Default speech-to-text, text-to-speech, and large language models are now included in the per-conversation price, with premium models available via Deepgram, ElevenLabs, and Anthropic.
What this means in practice
The platform aims to address the challenge of running AI agents reliably across customer journeys, providing visibility and control for enterprises. NiCE highlights a 50% success rate for AI autonomy, positioning Cognigy as a layer that ensures dependable AI performance rather than building frontier models. The new AI Assist tool simplifies agent creation for smaller businesses and contact center administrators, while advanced features cater to technical teams. The platform supports hybrid deployment models, allowing customers to ease into full agentic workflows without overhauling existing systems.
Limitations and availability
NiCE does not build frontier models, and the release does not include any new models beyond default and premium options. The platform supports hybrid deployment, with fully agentic use cases at the beginning of the adoption curve. Observability features like tracing and simulation are now part of the release, enabling data-backed model selection. Voice-to-voice support is available, though it is described as an early feature for customer deployment. The platform is designed for contact centers, with deterministic workflows remaining critical for compliance and security.
