NTT DATA Business Solutions
Simen Huuse | March 15, 2017

With SAP Cloud Platform, Rewarding the Usage of Chatbots Is Closer than You May Think


Ever since the popular TV show “Knight Rider,” people have been dreaming of dialoguing with machines. Now, the day has come when chatbots can be adapted for your business on a microcosmic or large-scale level, rather easily. This blog article will give you an idea of how chatbots can bring value to your everyday business – as both an external and an internal service.

Natural Language Processing

Chatbots can be built upon a combination of the SAP Cloud Platform and a provider of natural language processing (abbr. NLP), like the that originated in Palo Alto. NLP supports interpretation of the user’s intentions and passes that interpretation to the SAP Cloud Platform. Due to the bilateral connection between NLP and the SAP Cloud Platform, comprehensive information can be exchanged between the customer and the chatbot, whose information is sourced directly from your ERP.

Customer Service Driven by Chatbots

Imagine a scenario where a customer is surfing the website of a hardware store and a chat appears. If the customer asks: “Do you have a wheelbarrow available?” or “When can you have this wheelbarrow ready for pickup at the warehouse?” the NLP service would understand that the user is referring to the availability of a product at the warehouse. The term “wheelbarrow” would then be associated with a product categorized as a wheelbarrow, and once the intent and product were identified, a message would be passed to the SAP Cloud Platform. It would contain intent (stock information) and product type (wheelbarrow). The location could also be passed, if the device used could obtain the coordinates. Otherwise, the application would have to ask for it.

Within the SAP Cloud Platform, a Java Webhook (a kind of Java application) would choose what to do with the intent. In the example above, it would need to go into the SAP backend to get the relevant data. For this, the SAP Cloud Connector is the bridge between the cloud and the on-premise system. The stock information is fetched from the backend to the SAP Cloud Platform, and from there it is passed on directly to the customer. It should all happen so quickly that it feels like a normal conversation to the user.

The front end could be created with an SAP Fiori application, or it could be a 3rd party provider such as Skype, Slack, Cisco Spark, Amazon Alexa or Google Assistant. Such decisions should be based on the use case and the type of service the bot should provide.

IKEA is Ahead of the Curve with “Anna”

To learn more about chatbots, just keep an eye on IKEA. They already went live with an “Automated Online Assistant” called Anna. She answers questions like prices, sizes, delivery, spare parts and opening hours. She guides customers directly to the respective page they are searching for.

This is a great idea, because a bot is always available, and for those questions where it doesn’t have an answer, it can gather the necessary information for a customer-service representative to get back to the customer.

Chatbots Even Serve You Internally

But, let’s turn our perspective to internal processes. There are several use cases where chatbots could increase workflows. Let’s look at a factory, for example. On the shop floor, the use of mobile devices can be challenging, sometimes. Same goes for free text writing on a device. If the person interacting with the app could talk to a chatbot, this scenario would be more convenient than typing or searching for the right options. It would also reduce the time from when an issue is detected to when a maintenance order is created. Planning-process staff could record maintenance issues verbally and forward them via the ERP directly to the respective shift foreman of the maintenance team.

In addition to helping the inspector fill out a form, the assistant could ask questions based on the information previously provided to it, digging deep into certain areas that an inspector normally might overlook, or it could obtain information required by the maintenance worker who would do a repair. It could also suggest answers to common easy-solution problems. In this way, an assistant adds value far beyond form-based information gathering.

Technology Is Ready to Go

As you can see, there are lots of ideas going around. Hopefully, the use cases I tried to describe have inspired you with regard to your business. The technology is available to enable chatbot scenarios in order to let your business become more efficient or to boost its effectiveness.

Contact Us
Contact Us

Have questions? Please contact us.