AI Chatbots

AI chatbots can enhance the user experience on e-commerce platforms significantly. With the introduction of Chatgpt by ​OpenAI, numerous ​businesses spanning various industries have employed new generative AI chatbot experiences. As the AI market grows, many companies are expected to adopt AI chatbots. Studies conducted by Juniper indicate that global retail spending on chatbots may hit $12 billion by 2023 and potentially rise by as much as 470% to reach $72 billion within the next five years.

Given the trends in the AI market, it could be beneficial for even smaller businesses to consider devising and implementing modest AI chatbot experiences, both within the organization and for customer interactions. AI chatbots are predominantly utilized for customer support, with few businesses deploying these tools for product recommendations or to enhance the shopping experience.

Conversational AI

​Chatbots today often use Natural Language Processing (NLP) technology. This lets them understand human language by breaking it into smaller, manageable pieces. They can interact with and respond to both written and spoken words. An aspect of NLP is Natural Language Understanding (NLU), which improves how machines understand language. NLU uses different methods to find meaning, context, and insights from text or speech. It can also grasp different definitions of the same word depending on the context and correct language mistakes, like spelling errors.

Conversational AI combines NLP and NLU with traditional conversation tools like chatbots, voice assistants, and voice recognition systems. This can happen through speaking or writing. There are mainly three types of conversational AI:

Scripted Chatbots: The chatbots follow the rules and use decision trees to answer customer questions. Their set of responses is limited, and they’re often used to answer specific questions, like FAQs.

NLP Chatbots: NLP chatbots can provide more varied responses to customer questions because they can learn. They analyze the user’s input and determine its meaning to give the best answer or action.

Contextual Chatbots: These are the most advanced in processing language. They use AI and machine learning to remember past interactions and learn behavior patterns. They use NLU to understand language and detect feelings and intentions. They can also understand the context of a conversation and can correct language mistakes, making sure the conversation keeps going.

IBM Watson: AI Pioneer.

IBM Watson was one of the pioneers in the AI market. IBM Watson beat two of the most successful Jeopardy contestants twelve years ago. Proving that machines, when they are fed with large data sets, can understand human-produced questions. IBM developed a powerful AI system called DeepQA to help Watson win Jeopardy. DeepQA could understand and process the clues in the game and the information stored in Watson’s database. It took a team of 20 researchers three years to develop DeepQA, but it was worth it in the end, as Watson defeated its human opponents. Watson proved that machines could empower humans. It seemed like science fiction at the time. After Watson’s Jeopardy stunt, the company’s valuation soared.

The people at IBM were excited about what Watson could do for different industries, like law, healthcare, finance, and education. IBM’s marketing department had celebrities like Bob Dylan and Serena Williams talk to Watson to build trust with the public. Watson even went on the TV show 60 Minutes. People learned that Watson might be able to help with some of the world’s biggest problems, like climate change and cancer. For many people, Watson became a symbol of artificial intelligence. The buzz around Watson died down until ChatGPT brought AI to the public’s attention again.

IBM currently has over 100 million users of its AI systems, including high-profile companies such as Ernst and Young, SAP, General Motors, GSK, Samsung, Intel, and Moderna. In May this year, IBM launched a revamped platform called

​ is a comprehensive platform with three parts:

  1. A workspace for businesses to train and test AI models. It’s user-friendly, making it easier to build complex solutions using ​machine learning.
  2. ​ This toolkit helps businesses manage their data and AI effectively. It supports the creation of ​AI workflows that are responsible and clear. Companies can use it to handle more significant AI projects using all their available data. It also supports searches, management, and various data formats.
  3. ​Watsonx. governance: The platform allows companies to direct, manage, and monitor their organization’s AI activities, and it employs software automation.


Follow the step-by-step instructions below to start creating your IBM WatsonX assistant. Your bot should include at least three different actions, with more than three steps. Your bot should be deployed on your website.

IBM Discovery

After setting up your bot, follow the tutorial below to set up your Discovery Instance on the IBM Cloud Platform. This will allow us to train the bot using existing URLs and data. After setting up the discovery, add links to your different pages to the discovery, and add a document, or Excel file with FAQS.

Please find an additional tutorial explaining how to work with Watsonx