User Experience Insights
Building smart skills for chatbots – with SAP Conversational AI
Amazon’s Alexa, Microsoft’s Cortana, and Apple’s Siri have taken the world of cars, living rooms, kitchens, and bedrooms by storm. Thanks to the variety of available skills, these modern assistants can adapt continually to new fields of application and are establishing themselves as the natural communication channel between human and machine. Conversation with a non-human counterpart in real time has now become an accepted user interface for controlling devices, processes, and systems. As a result, intelligent chatbots and digital assistants are also finding their way into the business world. Any company planning to integrate AI must primarily be able to meet the wishes and needs of users interactively in every situation – regardless of whether they are using a smartphone, PC, or intelligent speaker; in essence: Conversational User Experience (CUX).
What chatbots and digital assistants should be able to do
So just what is Conversational User Experience (UX) and what makes it interesting for companies? A Conversational User Interface (UI) allows users to communicate with computers using natural language without having to learn a complicated syntax with specific commands. The task of a Conversational UX designer is to make these interactions a positive experience. Specific guidelines and tools for CUX support the design process. The objective is to ensure that users feel that their requests have been dealt with as well as if they were speaking to a human.
Two tools are available for intelligent text and voice conversation, These are, on one hand, chatbots for taking over everyday routines of many different users and, on the other hand, digital assistants that focus on providing targeted support for a specific user:
- Chatbots are software solutions that simulate human conversation. Natural language input, often in the form of simple text messages, is used to communicate with underlying systems. Chatbot systems can access an extensive repository of data and functions for selected task areas and provide decision-making support. Artificial intelligence is drawn on in this respect to enable answers to be improved continually.
- Digital assistants are a further step along in terms of development. Such a system combines many elements of chatbots, but is tailored additionally to ideally all everyday tasks performed by users, while also assisting users and learning from their behavior. In addition, an option is often provided to control digital assistants via voice and access services in this way.
Chatbots have already proven their worth in various practical applications, for example, for surveys or news in the area of marketing as well as for delivery status updates or recurring product queries in customer service. Intelligent chatbots can also be used to simplify use of complex systems. Note also though that it generally does not make sense to use this technology for specific queries, tasks that require transfer capacity, specific problems, or where feelings play a role during the course of conversation.
Developing a chatbot with SAP Conversational AI
SAP Conversational AI is a collaborative platform for creating chatbots with end-to-end integration. AI stands for Artificial Intelligence. Initially this technology was developed primarily for developing chatbots in B2C scenarios for customers, for example on websites. Meanwhile, this technology is also enhancing B2E scenarios (Business-to-Employee), through integration in applications from SAP S/4HANA or SAP SuccessFactors.
The platform supports collaboration for creating chatbots and enables content to be shared, which in turn can be used to create new scenarios. Content can be read dynamically from externally connected systems and displayed attractively in different ways. Data can also be generated after actions are initiated in a chat. Various messaging channels such as Facebook Messenger, Slack, and Microsoft Teams are supported for communication and visualization of chats, while integration in SAP Fiori apps is also possible. This means that SAP Conversational AI can already be used effectively today to develop chatbots that interact with external users or employees. Data from external sources and SAP data, for example from SAP C/4HANA, SAP SuccessFactors, or SAP S/4HANA, can be integrated easily and linked with customized logic.
Many scenarios can be implemented by simply using settings and scripting without the need for complex programming. The few technical hurdles that exist justify getting started straight away on projects for selected scenarios. In the same way as for traditional UX projects for PC or smartphone applications, a phase-based approach with a discover, design, develop, and deliver phase makes sense when it comes to Conversational UX projects too.
The current focus of the platform is on configuring own chatbots, which can also be used with non-SAP applications. SAP Conversational AI can be booked as a cloud service and connected with a customer’s own system landscape. The basic platform architecture consists of the following components:
It is especially important to remember that Conversational UX projects simply cannot be implemented without the involvement of users, because a machine learning model for conducting conversations has to be developed and trained in a training phase. This system needs conversations to learn which actions to apply in which situations. Test conversations with future users, which reflect the planned practical use, are therefore a critical factor for the quality of the application.
Key technology for interaction with users
By using artificial intelligence and machine learning techniques, modern chatbot systems are capable of automating tasks in a conversation process. This makes them accepted as a conversation partner. Various applications come into view for using this technology: from simple information on recurring questions through to booking scenarios with a range of input parameters. Yet chatbots can also be used for flexible retrieval of business data if the answer is not only to be provided in text form in the chat channel, rather also visualized by specific UI controls. Chatbot systems offer excellent potential for saving costs by transferring support tasks to bots. Improved service levels and location-independent anytime provision of services also speak in favor of their use.
Work is already under way on expanding SAP Conversational AI to a digital assistant for SAP systems or applications, which guides users individually through the daily work routine. By providing SAP Conversational AI as a cloud service, it is anticipated that the range of functions will grow at a rapid pace in short innovation cycles. Enhanced functions for working with business data or for collaboration in teams will then make in-house use even more interesting for companies.
All in all, the ongoing advances in the area of Conversational UX show that use of new and innovative technologies will continue to dramatically change how people work and use business applications. It is only recently that Conversational UX has even begun to play a role for most users of SAP applications. However, this type of communication could become a key technology for interaction with users in the coming years.
To find out the many other advantages for companies of using chatbots and how chatbots can be developed in practice, refer to the publication “User Experience mit SAP” (in German) for insights from the UX experts of the SAP partner itelligence.