Uniphore Help Center Portal

Uniphore's Agentic Systems are where you create and maintain all Administration Platform AI-based Agents.

Powered by Large Language Models (LLMs), AI Agents use natural language instructions to orchestrate the processes and take the actions needed to achieve a given objective. AI Agents autonomously analyze input, reason, and decide each step to take during a real-time session. They can evaluate the current conversation context, retrieve data from relevant business systems, policies or knowledge bases, and instantly respond in a natural manner, as well as perform any activities required to achieve session goals.

Each AI Agent is also assigned specific guardrails to prevent unwanted activities, hallucinations, and toxicity.

The scope of an AI Agent can vary greatly: from AI Agents that are able to achieve a broad range of tasks and objectives, to those that are built for a very focused objective. For example, a customer might design a single AI Agent that manages their entire contact center with a range of multiple tasks. Alternatively, the customer may create multiple simple AI Agents, each designed to be narrowly focused -- for example, an outbound lead prospecting. Large organizations might also create one set of AI Agents for their customers and a different set to assist their internal employees.