AI glossary for businesses

Chatbot

In short

A chatbot is software that holds a conversation with people through text or voice, usually on a website, in a messaging app or inside a company tool. Simple chatbots follow fixed menus and rules. AI chatbots use language models to understand free-form questions and write their own replies, which makes them more flexible but also less predictable.

How a chatbot works

Think of two kinds of receptionist. The first follows a laminated card: "for opening hours press 1, for orders press 2". That is a rule-based chatbot. It is predictable but gets stuck as soon as a question is not on the card. The second has read the company handbook and can answer in their own words. That is an AI chatbot, built on a large language model.

Many modern chatbots mix both. Fixed flows handle tasks like "track my order", where a wrong answer is costly. The AI part handles open questions, ideally answering only from your own documents and passing the customer to a person when unsure.

In practice for a small business

An e-shop selling clothing gets the same questions every day, often outside office hours: Where is my order? Can I exchange a size? Do you ship to Cyprus? An AI chatbot on the website, set up with tools like HubSpot's chatbot builder or Intercom, can:

  • answer policy questions in Greek and English from the shop's own help pages;

  • look up order status through a connection to the shop platform;

  • hand over to a person, with the chat history, when the customer is upset or the question is unusual.

Shops often find that a chatbot handles a good share of simple, repeat questions, but the actual share varies widely. Track it during a pilot. The AI for customer service replies page covers the setup.

Why it matters and what to watch out for

  • Wrong answers are your answers: if the bot promises a refund your policy does not allow, the customer will hold you to it. Limit it to approved content and test it with real questions.

  • Tell people it is AI: the EU AI Act requires that people are informed when they interact with an AI system, unless it is obvious. Check the details with your legal advisor.

  • Easy exit: always offer a clear way to reach a person. Customers stuck in a loop leave.

  • Chat logs are personal data: names, addresses and order details in chats fall under GDPR. Set retention periods and check where the vendor stores data.

  • Maintenance: update the bot's sources whenever prices, policies or delivery times change.

Want an AI roadmap like this for your business?

In a two-day workshop, we turn your team’s ideas into a prioritized AI plan you can start executing within 90 days.

Related terms