SAP Conversational AI and NLSQL are two good tools to create compelling cloud-based services and solutions. The recent availability in both platforms of AI capabilities has paved the way for a new class of innovative and autonomous enterprise applications. Customer support ChatBots are one of these kinds of applications, and SAP Conversational AI product in this field comes in the form of a web-based graphic interface for developing chatbots with dialog scenarios based on Recast.AI French startup, which was acquired a year ago by SAP. NLSQL has a broad range of services, including unstructured language conversion to structured quires API, natural language interface to SAP, data visualization and the most relevant questions AI prediction during first chat user engagement.
This article shows how powerful it is to combine NLSQL with SAP Conversational AI services to build a chatbot service desk that interacts with NLSQL to log and retrieve data from enterprise systems.
SAP Conversational AI includes:
– Prebuild skills: such as Greetings, Small Talk, Weather, Customer Satisfaction, Set Alarm
– Possibility to build any conversational flows with buttons without programming skills
– Integration into the most popular messenger platforms Slack, MS Teams, Skype, Skype for Business, Facebook massenger, Telegram, etc.
– Integration with fallback channels as SAP Contact Center or Intercom for human interaction to conversational flow in case of need
– Possibility for integration with 3rd API web-services
Actually, it is all the required tools for creating AI conversational customer support chatbot. Using the chatbot customers can follow prebuild dialog flows or chat with human support in case of fallback occurs. But there is no easy way for integration to company systems or databases itself because systems couldn’t deal with unstructured natural language requests from users. APIs or databases require Structured Query requests only for providing any calculated information to the user.
Integration with NLSQL provides the possibility for transforming users unstructured natural language requests into structured quires by calling external API web-service. Basically, it is an external API, which helps people to talk with machines using structured query language. Translation from human language to machine language can help people to get numbers from CRM or ERP systems or quire the databases for required information.
So, on the picture above you can find all components are put together in a combined solution.
For the solution implementation it is highly recommended to check the following tutorials:
Basic steps for creating the chatbot with a feature to present data from the database are described into above-mentioned tutorials, — first is for the chatbot application setup and second is for integration with NLP to SQL API
That’s it! Once steps are done your chatbot solution is ready and your first call for data can be made.