Agent Assist
Agent Assist helps helpdesk agents respond faster and more accurately by analyzing both written and real-time voice interactions. It can use large language models (LLMs) to understand client messages and suggest the most relevant replies from your Knowledge Base.
Agents can also ask Agent Assist specific questions to get quick, AI-powered answers.
How It Works
When a client-bot conversation is transferred to a live agent in DRUID Live Chat, Agent Assist activates automatically—if the Agent Assist special flow is configured. It processes each message in real time as follows:
- Captures the full conversation transcript by updating [[ConversationInfo]].AssistConversationItem[i], storing:
- The message in [[ConversationInfo]].AssistConversationItem.Message
- The message originator in [[ConversationInfo]].AssistConversationItem.Originator (user for client messages, agent for helpdesk agent messages).
- Searches the Knowledge Base using each client message to find relevant content.
- Triggers the Agent Assist special flow to generate a response suggestion based on the client’s message.
- Sends the suggested reply to the helpdesk agent.
- Prevents overlapping suggestions by ensuring only the latest client message triggers a response—after the previous Agent Assist flow has completed.
Set up Agent Assist for response recommendations
- Open Solution Library, search for solution Agent Assist + Live Chat, and import it.
- Go to Bot Details > General details > Dialogue Management.
- Select Use Knowledge Base.
- From the Knowledge Base response flow field, select Agent assist flow.
- Go to Apps and configure the connection strings for the GPT-azure.com app.
The solution comes by default with Druid Data Service integration, saving the suggested responses to the Agent Assist workspace that comes with the solution.
After setup, Agent Assist will automatically suggest responses, helping agents provide faster and more accurate support.