What Is a Chatbot? Definition, Types, and Examples

You have most likely encountered chatbots in customer service, when you need help accessing your bank account, returning a pair of shoes, booking an appointment, or troubleshooting software on your computer. Some chatbots are now integrating with artificial intelligence (AI) to deliver personalized assistance. 

Learn what a chatbot is, types of chatbots, how they work, and several examples of chatbots. If you want to learn more about chatbots, and how to build them, you’ll also find courses on chatbot development at the end of this article.

What is a chatbot?

Chatbots are computer programs that simulate human conversation, written or spoken. These days, chatbots are starting to integrate conversational AI, such as natural language processing (NLP), to understand questions even if it isn’t grammatically correct and then respond based on data it has collected.

A chatbot may prompt you to ask a question or describe a problem, to which it will either clarify what you said or provide a response. Some are simple, giving a response only to the question asked. Some are sophisticated, learning information about you based on data collected and evolving to assist you better over time.

Chatbots vs. AI chatbots vs. virtual agents vs. virtual assistants

How do chatbots work?

Chatbots process collected data and often are trained on that data using AI and machine learning (ML), NLP, and rules defined by the developer. This allows the chatbot to provide accurate and efficient responses to all requests. The two main types of chatbots are declarative chatbots and predictive chatbots.

Types of chatbots

Chatbot type is determined by its capabilities. Declarative chatbots are more basic than predictive chatbots.

Declarative chatbots perform one function. These chatbots use NLP, defined rules, and ML to generate automated responses when you ask a question. Declarative, or task-oriented chatbots, are most common in customer support and service–and are best when answering commonly-asked questions like what the store hours are and what item you’re returning. This type of chatbot is common, but its capabilities are a little basic compared to predictive chatbots.

Predictive chatbots are more sophisticated and personalized than declarative chatbots. Often considered conversational chatbots, or virtual agents, these AI- and data-driven chatbots are much more interactive and aware. They utilize NLP and more complicated ML, along with natural language understanding (NLU) to continue learning about the user through predictive analytics and intelligence. Over time, they can even predict recommendations and anticipate your needs.

Guided Project: Create your own chatbot

Create your own chatbot in this two-hour project. You’ll use Rasa, a framework for developing AI-powered chatbots, and Python programming language, to create a simple chatbot. This project is ideal for programmers who want to get started in chatbot development. 

Who builds chatbots?

Chatbots tend to be built by chatbot developers, but not without a team of machine learning and AI engineers, and experts in NLP. Here are a few careers involved in building chatbots.

Chatbot developer: These professionals build a conversational experience for users with AI, machine learning, and natural language processing.

AI engineer: AI engineers build models using machine learning algorithms and deep learning neural networks that can be used to make decisions, such as in the production of chatbots.

NLP engineer: NLP engineers create programs that can understand human languages and respond accordingly, using a combination of computer science and AI.

Examples of chatbot use cases

Chatbots are useful in a great variety of ways. Here are a few common examples.

Engaging with a mobile app

You might use a chatbot in a mobile app when you’re paying for an item or subscription. It might offer the option of direct monthly payments from your bank instead of manually paying each time. In a doctor’s office, you might fill out intake forms on your phone with the help of a chatbot.

24/7 customer service support

It is common for customers to consult a chatbot when they’re having trouble logging into their account, or when they have an issue with a tool or system software and want to ask questions (without calling customer service). 

Personalized recommendations

AI chatbots like Alexa are continuously learning from data they collect, so you may ask a chatbot whether it is raining outside–and it might offer personalized recommendations like setting an earlier alarm to account for traffic on your morning commute.

Build AI-powered chatbots with IBM

Learn how to create a chatbot without writing any code, and then improve your chatbot by specifying behavior and tone. Deploy the chatbot on a WordPress website and interact with it. Do all this and more when you enroll in IBM’s 12-hour Building AI Powered Chatbots class.