Building a Chatbot Application with IBM Watson


The market for chatbots is expanding and is expected to continue growing. Did you know that IBM Watson can help build intelligent and efficient AI-based chatbot solutions?

AI-powered chatbots provide a more human-like experience, can carry on a natural conversation, and continuously improve over time. Their popularity stems from the improvements they bring to customer experiences, their 24/7 availability, cost-effective customer support, enhanced operational efficiency, easy scalability, support for multiple channels, data collection and analysis capabilities, continuous learning, and task automation. These benefits make them valuable tools for businesses across various industries. The working of an AI-based chatbot is demonstrated in Figure 1. As shown, the chatbot evaluates the users’ requests, ascertains their intent, determines what they require, and then automatically generates and replies to them with the most relevant response.

How AI-based chatbots work
Figure 1: How AI-based chatbots work

Top industries using AI-based chatbots for their business

Various industries widely adopt AI-based chatbots to enhance customer engagement, improve operational efficiency, and provide personalised assistance. For example, chatbots are utilised in the travel industry for tasks like booking flights, hotels, or rental cars. They can also offer travel recommendations and information about destinations and answer frequently asked questions. Chatbots can assist in itinerary management and provide real-time updates about travel schedules. In the insurance industry, chatbots help customers file claims, obtain policy information, and answer questions about coverage. They can guide users through claims, collect relevant information, and offer instant assistance. Table 1 shows examples of how different industries are leveraging AI-based chatbots for their businesses.

Table 1: Popular AI-based chatbots used in industries

Company name AI-based chatbots used for their business
Bizbike is Belgium’s biggest e-bike provider. It has implemented AI chatbots to efficiently and reliably respond to consumer enquiries.
Starbucks uses AI chatbots to place orders via voice and carry out payment transactions. It also uses them for digital marketing and retail purchase.
Belfius Belfius is a popular, trusted bank in Belgium that uses AI chatbots to manage insurance claims and respond to customer queries efficiently.
iFood iFood is a Brazil-based company that uses an AI chatbot to manage an online meal delivery system.
European Commission The European Commission uses an AI chatbot to manage traveller’s queries.
Kotak Life Insurance India’s Kotak Life Insurance has developed the Kaya virtual assistant as an AI-based chatbot system to manage user queries.

Several AI-based chatbot tools are available in the market, each with its own features and capabilities (Table 2).

Criteria WP-Chatbot Microsoft Bot IBM Watson Assistant CSML LivePerson
Developed by MobileMonkey Microsoft IBM LivePerson, Inc.
AI technology used NLP NLP NLP, ML NLP, ML NLP
Key features Facebook page branding, compatible with WordPress Allows use of open source SDK Provides a visual interface for designing chatbot flows, and supports multiple channels Provides a ready-to-use chatbot templates library Hyperlinks and custom pre-written statements
Target application
WordPress ecosystem To build a conversational AI experience To respond to user queries for commerce and support To improve the customer experience for e-commerce platforms To engage customers using messaging, voice, etc
Pricing 14-day free trial; and US$ 99/month No free trial
US$ 499/month
US$ 140/instance Free version available Free trial available

Using IBM Watson to develop a chatbot application

The steps we need to follow when using IBM Watson to deploy an AI-based chatbot solution are briefly listed below.

Step 1. Set up an IBM Cloud account: If you don’t already have an IBM Cloud account, sign up at

Step 2. Create an IBM Watson Assistant service:  Once you successfully log in at IBM Watson, create an assistant service using the catalogue menu bar. Now, generate assistant and dialogue skills. Give a name to your dialogue skill, for example, ‘student advisor’, as shown in Figure 2. An assistant is a chatbot that helps connect user conversation flow and search for an answer from the knowledge database.

 Create a chatbot assistant to address customer queries
Figure 2: Create a chatbot assistant to address customer queries

Step 3. Design the conversation flow: Use the Watson Assistant tooling to design the conversation flow by creating nodes and defining the dialogue structure. Each node represents a step in the conversation, including user inputs and chatbot responses.

Step 4. Define intents and entities: Identify the intents (user goals or intentions) and entities (important pieces of information) that your chatbot should understand. Train Watson Assistant to recognise these intents and entities to improve the accuracy of understanding user inputs. You can construct an intent called ‘feedback’ that acts in response to user input, as shown in Figure 3.

Create intent to define the user queries
Figure 3: Create intent to define the user queries

Step 5. Add a custom greeting node: Click the ‘Add node’ button in the workspace editor to create a new node, as depicted in Figure 4. This node will represent the custom greeting.

Adding a new custom greeting node
Figure 4: Adding a new custom greeting node

Step 6. Save and test: Finally, save your changes and try the custom greeting node by interacting with the chatbot to ensure it triggers the greeting when the specified input is provided. A sample deployed application is shown in Figure 5.

Test the greetings chatbot you developed
Figure 5: Test the greetings chatbot you developed

Chatbots are gaining popularity in industries with their ability to enhance customer experiences, improve operational efficiency, reduce costs, and provide businesses with valuable insights for better decision-making. IBM Watson is a powerful platform that offers a range of AI capabilities, including natural language processing (NLP) and machine learning, which can be leveraged to build AI chatbots. Businesses can build robust and intelligent chatbot solutions by utilising IBM Watson’s AI capabilities.


Please enter your comment!
Please enter your name here