Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
cancel
Showing results for 
Search instead for 
Did you mean: 
former_member363931
Participant
980
Are you using a chatbot to relieve your customer service and reduce the average waiting time for your customers? Scared that your chatbot may not be able to meet all your customers' needs, and that it may cause a worse experience in certain areas?

In this article I would like to show you how you can work around this problem using the examples of SAP Conversational AI and SAP Customer Engagement Center.

If you do not yet use a chatbot in your company, you may be interested in this post, which explains why you should also use a chatbot.

Chatbot Architecture - The Receptionist Pattern


You'll probably already know what to consider when building your own chatbot. As an architecture pattern the colleagues from SAP Conversational AI suggest the Receptionist Pattern.

The Receptionist Pattern is a bot design pattern that positions the bot at the beginning of each individual user request. Each question is addressed to the bot, who should then be able to understand what the request refers to.

Thus, the core is that the chatbot understands every single request. However, this does not mean that the chatbot has to handle all requests autonomously. After understanding the question, the bot can either:

1. process the request autonomously.

2. start the conversation to collect important information, e.g. customer number, e-mail address, and then hand the conversation over to a human agent.

3. hand over the request directly to the right group of human agents, depending on the topic.



In the following, we will concentrate on variants 2 and 3 and take a look at how the SAP Customer Engagement Center can support us here.

The desktop for your agents - SAP Customer Engagement Center


The SAP Customer Engagement Center is part of SAP C/4HANA, more precisely the SAP Service Cloud. As a contact center or call center, it stands between self-service solutions such as SAP Conversational AI and SAP Multichannel Foundation for Utilities, and field service solutions such as SAP Fieldservice Management.

It offers a unified fiori-based agent desktop with integrated communication channels, ticketing and insight into customer data and interaction history.

SAP wants to simplify customer service across all channels with the SAP Customer Engagement Center (CEC) by providing the ability to manage all customer interactions for a unified service experience in the same interface.

The fallback - your agent takes care of the customer


In SAP Conversational AI, you can define situations in which your chatbot should transfer the conversation to a fallback channel. Here, for example, your call center agents can take over the conversation.

The possible fallback channels for your chatbot can be configured in the botbuilder under the tab "Connect":



The SAP Customer Engagement Center is one of the possible solutions. The description of the setup on SAP Conversational AI side can also be found there.

After setting up the fallback channel you can use it in your skills. 

For example, in the general fallback skill. For instance, you might want to ask a user if he wants to talk to a human agent at the third time the fallback is triggered.



On the other hand, you can also pass on to human agents within special skills. For example, you could cover topics that are of interest to your target group, but which you cannot yet process automatically.



For your agent, it is a normal chat conversation that he receives via the corresponding queue. However, he also sees the customer's previous conversation with your chatbot and can thus, for example, obtain valuable information from the context of the conversation or the queries of your chatbot and reduce the processing time.

After the agent has ended the conversation, the customer can continue interacting with the chatbot.

In combination, the strength of the two solutions can be seen. Your chatbot can receive customer requests and possibly process them without the involvement of an agent. If your chatbot cannot offer a solution for your customer, or if the customer wants to talk to a human, an agent can help him.

The result - always the best experience for your customers!


For you there are several advantages to combining the solutions. You don't have to worry that your chatbot isn't perfect. You won't be able to guess all your customers' questions in advance. But you don't have to, because you can rely on your agents in any case of necessity.

On the other hand, your customers will be able to find solutions faster. They can get quick solutions through your chatbot at their own pace and time. At the same time they also have the opportunity to write with a human during your service hours.



From my point of view this is a meaningful measure to increase the acceptance and the satisfaction with a chatbot. What do you think?

Best regards,

Tobias

P.s. A German version is available in a guest article here.