Agent Assist improves productivity by suggesting the next best action (Bots or Knowledge Base articles) to agents in real time, based on consumer’s intent.
The Conversation Orchestrator analyzes all available automations for the identified intent and recommends the best one in real time to the agent. Recommendations are done by choosing the highest ranked bot or KB article by score, after calibrating the score using a machine learned signal about the historical performance. Agents can opt to manually use the KB article to help the user, or join the bot to the conversation, monitor, and remove if needed.
For more information on implementation, please see the LivePerson Developer Center.