Building HR Chatbots with SAP Conversational AI [Artificial Intelligence]


Artificial intelligence has enabled us to ensure that repetitive and mundane tasks are handled by intelligent chatbots that can do those activities for you or your department. Often, these chatbots handle traffic that could be overwhelming and highly unproductive for the team. The delegation of such tasks to an AI chatbot built with a low code high efficiency platform makes things extremely worthwhile and interesting. This will be part one of a three part series on how to use SAP conversational AI to build chatbots for HR purposes.

Concept of a HR Chatbot

An HR chatbot has to primarily aim to take off the load from the HR team members vis-à-vis repetitive and non productive tasks such as employee queries. The reason the chatbot would be highly effective is that a) the employee experience would be retained as a 1 to 1 interaction with the HR department through this chatbot and b) it would free up time for the HR team to concentrate on their core duties.

SAP Conversational AI BOT Builder

The three main parts of a BOT builder process are:

a) Getting – Firstly we need to get the input from the user to actually begin doing what the BOT is supposed to do. This part deals with how we intend to get that information from the user. This is where the BOT connector comes in. The BOT connector basically helps in catching the incoming message and then it sends this message to the BOT builder.

b) Understanding – Understanding what a user has requested for is a challenge of its own. Every Chatbot has to ensure that the instruction/ request has been well understood before it can start to process it. Natural language processing comes in at this juncture to save the day! Yes, once the information or instruction is received from the user by the BOT builder, it uses the NLP API to decode information from it.

c) Conversation – This is the final point where the BOT interacts with the user using its skills and the conditions that it has been fed.

Skills and Conditions

Skills are basically used to group together a specific set of instructions that will help your chatbot perform those steps and achieve a particular end. For e.g. if my chatbot has to book a flight for my employee upon taking inputs of start date, end date, destination etc. I would add a skill called book-flight that would achieve this goal.

Now the chatbot cannot just go around performing activities without a certain condition being met – else it will just be looping over. Therefore, conditions!

Conditions consist of triggers, requirements and actions.

Triggers – These basically tell the bot when a particular skill is to be executed. It mentions for the BOT to check if a particular skill is present or not and thereby take a decision on the selection.

Requirements – These are basically the inputs that are retrieved by the BOT before it can proceed to take “actions”. Each requirement is nothing but an input that the BOT requires before it can make a decision or, more formally, an action. These requirements are to ensure that the skill can be executed after all the necessary variables from the user are received.

Actions involve various outcomes that a BOT can generate and they are of different categories as below:


SAP Conversational AI platform enables chatbot building and makes it a task that is devoid of unnecessary coding complexities. Once we master the various building blocks of this platform, it is quite rewarding to get onto chatbot building and enabling better employee and HR experience.