MAS: Optimizing Contact Centers Like Never Before

Contact center managers are always on the lookout for cutting-edge solutions to enhance customer experiences and boost operational efficiency. Enter Multiagent Systems (MAS)—a groundbreaking technology composed of multiple independent but interactive Generative AI (GenAI) Agents. These systems offer a robust and adaptable approach to delivering exceptional value in complex CX environments. In this blog, we'll dive into a real-life example of how MAS can transform customer interactions and revolutionize contact center operations. 

Example of a MAS in a Contact Center 

Let's explore a typical customer interaction in a contact center and see how MAS can elevate the customer experience and boost productivity by automating tasks and supporting human agents. 

Imagine I'm a customer seeking advice from a bank about a new mortgage. Traditionally, the call flow would look something like this when a customer calls the organization: 

Traditional Call Flow: 

1. Caller navigates through an IVR menu and is routed to a human agent. 
2. The human agent assists the customer with their inquiry, supported by a CRM system and/or other applications and data sources. 
3. The agent updates the CRM records after the call is finished in wrap-up time. 

Enhanced Call Flow with MAS: 

1. Call Initiation and Routing Stage: 

The Orchestration Agent listens in real-time throughout the call and can be seen as the ‘project manager’ in the MAS during the call process. So when a new call arrives, the Orchestration Agent gets activated and directs the call to an AI Agent responsible for welcoming the customer and determining the call's context. Utilizing advanced Natural Language Understanding and Speech software, this feels like a normal conversation to the caller without any need for menu structures or DTMF inputs. Recognizing the context of the call as a financial inquiry, this Agent involves a second AI Agent to conduct an identity verification process with the caller. Once verified, the call is intelligently routed to a suitable human financial advisor. 

2. Human Agent Handling the Call with Knowledge Assistance: 

Meanwhile, the Orchestration Agent, now informed of the caller's identity, activates the Knowledge Assist AI Agent. This agent gathers initial customer details from the CRM and other sources, passing this information to the human agent when the call is presented. The Knowledge Assist Agent continues to listen in and keeps providing the human agent with relevant information and recommendations from various data sources based on the context of the call. This ensures the human agent has all the necessary information at hand, without needing to spend time searching.

3. Call Finish and Afterwork: 

After providing the customer with the best possible mortgage advice, the call concludes. At this point, the Orchestration Agent activates a Summary Agent, which instantly creates a summary of the dialogue and presents this to the human agent. With a single mouse click, the human agent can approve this summary, and the AI Agent ensures all relevant CRM records are updated. If needed, the Orchestration Agent can also schedule follow-up activities before completing the call flow.

Figure 1: Deployment of MAS to support human agents in a contact center 

Benefits of Deploying MAS in Contact Centers 

Throughout the customer call handling process, multiple AI Agents support the transaction, automating tasks and empowering the human agent to be more knowledgeable and efficient in helping customers. Deploying MAS in this scenario results in a better, more consistent customer experience while also optimizing productivity by reducing the time and effort required from the human agent. 

Real-Life Customer Success Story: Rabobank Netherlands 

Rabobank, one of the largest financial institutions in the Netherlands, is a shining example of how MAS can drive productivity, increase agent happiness, and elevate customer satisfaction. Partnering with Anywhere365 and utilising their Deepdesk AI platform, Rabobank implemented a multiagent system to support their agents. The Knowledge Assist feature connects all the disparate knowledge sources their agents rely on (FAQs, knowledge bases, internal wikis, etc.) and, based on the voice conversation, automatically provides the agent with the information they need. 

Imagine a customer calling to replace a lost credit card. Deepdesk MAS instantly supplies the agent with all the necessary details to assist the customer efficiently. After the call, the Summarizer generates a summary for the agent to add to the customer's CRM record. This seamless integration of AI support ensures that agents have everything they need at their fingertips, enhancing both productivity and customer satisfaction. 

Conclusion 

The future of Multiagent Systems (MAS) is incredibly exciting. From revolutionizing customer interactions to optimizing entire business ecosystems, the potential of MAS is boundless. MAS will transform the contact center industry and optimise customer service for customers whilst improving agent experience and efficiencies. 

Stay tuned for our next blog to discuss how you could start with the deployment of this transformative technology in your organization.