Genesys Agent Copilot

Agents balance many tasks simultaneously: speaking to customers (sometimes more than one at a time), reviewing data, finding answers, planning what to say and do next, and capturing notes. As consumer preference and increasingly advanced self-service options move “easy” questions out of the contact center, agents are left with complex issues to solve, for customers who have expectations that are higher than ever before.  

To deliver standout customer experiences, agents must have useful data and insights, within the work space that they’re already using, at the moment of need.  

Genesys Cloud Agent Copilot empowers contact center agents with AI-driven guidance during and after customer interactions. Genesys Cloud Agent Copilot determines customer intent, automatically surfaces knowledge and guides agents to their next best actions, summarizes interactions, and predicts wrap-up codes. 

Positive customer experiences can only happen when agents can answer a customer’s request, provide empathetic, personalized service, and deliver on the requested outcome. In many contact centers, agents must navigate multiple processes and tools to look up knowledge and FAQs to find answers and resolve customer inquiries which takes time, leaves customers sitting on hold, and causes high average handle time and long waits.

With Agent Copilot, companies can leverage the power of artificial intelligence (AI) to support agents as they serve customers digitally, or on the phone. Genesys Cloud Agent Copilot determines customer intent, automatically surfaces relevant knowledge, and guides agents to their next best actions – such as what to say next, what workflow to kick off, how to follow up, and more. It then summarizes interactions by generating text and predicts wrap-up codes. Agents spend time personalizing the assistance they offer each customer, and providing superior answers based on  suggested results, rather than digging for information and writing call notes.

It is possible to create rules based on events like starting an interaction, ending a conversation, or transferring it to another queue or agent. When an event like this happens, it will be possible to configure the appearance of an article from the knowledge base, triggering a script or a canned response.

Agent Copilot also allows to train a NLU model that detects Intents based on utterance. The detection of these intents will allow the user get an article from the knowledge base, triggering a script or a canned response, as with the events described above.

After the completion of the conversation between the agent and the user, a summary of the conversation is generated. This summary can be reviewed, modified, copied and pasted as part of the notes of the interaction.

Genesys offers Agent Copilot as a native AI capability fully integrated into Genesys Cloud CX.

During customer interactions, Genesys Cloud Agent Copilot assists agents by presenting relevant, next best actions to the agent in their workspace. Agent Copilot understands customers’ intents and provides agents with relevant answer highlighted knowledge, canned responses, and scripts. When agents move to after call work, Genesys Cloud AI creates an interaction summary and predicts wrap codes.

Business Logic

BL1: Review knowledge: The agent performs a high-level assessment to ensure the information returned from Agent AssistAgent Copilot is appropriate and relevant to the current conversation. 

BL2: Leverage suggestion: The agent communicates relevant information to the end-customer, or they use the information to perform the required “back-end” actions to resolve the customer issue. 

BL3: Rate suggestions: Agent Copilot may provide an agent with multiple pieces of information during the interaction. Agents should rate the information using the thumbs up / thumbs down buttons to verify as relevant or irrelevant. 

BL4: Resolve issue or continue conversation: If the end-customer issue is not adequately resolved, the agent continues the conversation with the end-customer to trigger Agent Copilot to surface additional information. If Agent Copilot is unable to provide appropriate information to resolve the end-customers issue, Agents should follow their corporate escalation policy to ensure that expectations are fulfilled. 

BL5: End Coversation: Agent reviews the AI generated note and Wrap Up Codes. Agent selects the wrap up code from the AI suggested or finds the appropriate code.  

Distribution Logic

Since the end-customer is already speaking with an agent in real time, any subsequent call steering is likely to be manually directed by the agent. 

N/A

Real Time Reporting

The Agent Copilot Performance dashboard for Genesys Agent Copilot gives overview about knowledge base article activities. Genesys Agent Copilot metrics and reporting provides insight about presented, opened and copied articles. For more information, see https://help.mypurecloud.com/?p=280180  

Historical Reporting

In the knowledge optimizer dashboard, you can analyze the effectiveness of your knowledge base. In this view, you can see the following metrics: 

• All queries in a specific time frame and the breakdown, in percentages, of answered and unanswered queries. 

• All answered queries in a specific time frame and the breakdown, in percentages, of the application from which the conversation originated. 

• All unanswered queries in a specific time frame and the breakdown, in percentages, of the application from which the conversation originated. 

• Top 20 articles and the frequency in which an article appeared in a conversation. 

• Top 20 answered queries and the frequency in which each answered query appeared in a conversation. 

• Top 20 unanswered queries and the frequency in which each unanswered query appeared in a conversation. 

see https://help.mypurecloud.com/articles/knowledge-optimizer-overview/ 

Customers and/or Genesys Professional Services are responsible for managing the Copilot NLU, rules engine and uploading their own knowledge base content into Genesys Knowledge Workbench to be used by Agent Copilot. 

Customer needs to provide a KB or the articles that will be the elements of the Knowledge Base.

V 1.0.0