Reference
AI Chatbot Glossary
56 essential terms for understanding AI chatbots, conversational AI, and customer automation — with clear definitions, practical context, and links to product pages where each concept comes to life.
AI Agent
An AI agent is an autonomous software program that perceives its environment, makes decisions, and takes actions to achieve goals — going beyond simple chatbots by handling multi-step tasks.
AI Chatbot
An AI chatbot is a conversational interface powered by artificial intelligence that understands natural language, generates human-like responses, and assists users with tasks like answering questions, capturing leads, or processing orders.
API (Application Programming Interface)
An API is a set of protocols that lets different software applications exchange data. In chatbots, APIs connect the bot to CRMs, e-commerce platforms, and helpdesks so conversations trigger real-world actions.
Auto-Reply
An auto-reply is an automated message sent by a chatbot or messaging system without human intervention — triggered by keywords, user actions, or time-based rules.
Bot
A bot is a software application that runs automated tasks. In customer communication, bots handle repetitive interactions like answering FAQs, routing inquiries, and collecting contact details.
Bot Training
Bot training is the process of providing a chatbot with data, instructions, and examples so it can respond accurately to user queries.
Bounce Rate
Bounce rate is the percentage of website visitors who leave a page without interacting or navigating further.
Chatbot
A chatbot is a software application that simulates human conversation through text or voice — ranging from simple rule-based systems to advanced AI-powered agents.
Chatbot Widget
A chatbot widget is an embeddable UI component that adds a chat interface to any website — typically a floating button that opens into a conversation window.
Conversational AI
Conversational AI is a branch of AI focused on enabling machines to understand, process, and respond to human language in a natural, contextual way.
Conversion Rate
Conversion rate is the percentage of visitors who complete a desired action, such as making a purchase or submitting a lead form.
CRM (Customer Relationship Management)
A CRM is a system for managing all your company's interactions with current and potential customers — chatbots sync captured leads and conversation history into the CRM automatically.
Customer Support Automation
Customer support automation uses technology — including AI chatbots — to handle customer service inquiries without human intervention.
Embedding
An embedding is a numerical representation of text in a high-dimensional vector space — used to measure semantic similarity between pieces of text.
Exit Intent
Exit intent detects when a visitor is about to leave the page — chatbots use it to display a targeted message or offer before the visitor bounces.
Generative AI
Generative AI is a category of AI that creates new content — text, images, code — instead of just classifying or predicting from existing data.
GPT (Generative Pre-trained Transformer)
GPT is a family of large language models developed by OpenAI that can understand and generate human-like text. Chatonbo supports GPT-5.4 as one of its model options.
Hallucination
A hallucination is when an AI model generates information that sounds plausible but is factually incorrect — a key risk in generative AI chatbots, mitigated by RAG and strict prompts.
Human Handoff
Human handoff is the process of transferring a chatbot conversation to a live agent when the bot cannot resolve an issue or the user requests human help — preserving conversation context.
Intent Recognition
Intent recognition is the AI capability of identifying what a user wants to accomplish from their message — critical for routing conversations correctly.
Integration
An integration is a connection between your chatbot and an external tool — CRMs, helpdesks, messaging apps, e-commerce — enabling data sharing and automated workflows.
JavaScript SDK
The JavaScript SDK is the lightweight embed script that loads a chatbot widget on a website — typically a single <script> tag with a bot ID.
JSON Payload
A JSON payload is the structured message a webhook sends to an external system when an event happens in the chatbot, such as a new lead being captured.
Large Language Model (LLM)
An LLM is an AI model trained on vast amounts of text that can understand and generate human language — LLMs like GPT-5.4, Claude, and Gemini power modern chatbots.
Lead Capture
Lead capture is the process of collecting contact information (name, email, phone) from website visitors during a chatbot conversation — weaved naturally into the dialogue.
Lead Qualification
Lead qualification is the process of evaluating whether a potential customer is a good fit — AI chatbots qualify by asking targeted budget, need, timeline, and authority questions.
Live Chat
Live chat is a real-time communication channel connecting website visitors to human agents — often combined with AI chatbots that handle initial inquiries.
Machine Learning
Machine learning is a branch of AI where systems learn and improve from data without being explicitly programmed — the basis for modern chatbot language understanding.
Multi-Language Support
Multi-language support is a chatbot's ability to understand and respond in multiple languages — auto-detecting visitor language and replying fluently.
Multilingual Chatbot
A multilingual chatbot can converse fluently in two or more languages, auto-detecting the visitor's language and replying accordingly — eliminating language barriers.
Natural Language Processing (NLP)
NLP is a field of AI focused on enabling computers to understand, interpret, and generate human language — powering chatbot capabilities like intent recognition and entity extraction.
Natural Language Understanding (NLU)
NLU is a subset of NLP focused on machine reading comprehension — understanding meaning behind user messages, not just keyword matching.
Omnichannel
Omnichannel is a customer engagement strategy providing a seamless experience across multiple channels — website chat, WhatsApp, email, SMS — with context preserved between them.
Open-Ended Conversation
An open-ended conversation lets users type anything in their own words rather than selecting from predefined options — AI chatbots excel because they understand free-form input.
Personalization
Personalization tailors chatbot responses to individual users based on browsing history, past purchases, location, or prior conversations — driving engagement and conversion.
Prompt Engineering
Prompt engineering is the art of crafting instructions that guide AI models to produce desired outputs — defining chatbot personality, knowledge boundaries, and response rules.
Quick Replies
Quick replies are pre-set buttons inside the chat window letting visitors respond with one click — speeds up common questions and reduces drop-off.
Query Understanding
Query understanding is the AI step of interpreting what a user is asking — even when vague, misspelled, or phrased with slang — via semantic search plus language models.
RAG (Retrieval-Augmented Generation)
RAG is a technique that enhances AI responses by retrieving relevant information from a knowledge base before generating an answer — reducing hallucinations and grounding replies in real data.
Response Time
Response time is how long a chatbot takes to reply — typically 1–5 seconds for AI chatbots, much faster than human agents, which lifts customer satisfaction and engagement.
Sentiment Analysis
Sentiment analysis is the AI capability of detecting emotion in user messages — happiness, frustration, confusion — and triggering handoff or prioritization based on it.
Session
A session is a single continuous interaction between a user and a chatbot — tracking conversation history, user context, and captured data from first message to end.
Smart Reply
Smart replies are AI-generated response suggestions that appear as clickable options during a conversation — anticipating the user's next message to speed up interactions.
Token
A token is the basic unit of text an AI language model processes — roughly a word or sub-word. Token counts affect API costs and the max conversation length a model can handle.
Training Data
Training data is the information used to teach an AI model how to respond — for chatbots, this includes knowledge base documents, website content, FAQs, and product catalogs.
Trigger
A trigger is a condition or event that causes a chatbot to take an action — page load, time on page, scroll depth, exit intent, or specific URL visit.
Webhook
A webhook is an HTTP callback that sends real-time data from one application to another when an event occurs — used by chatbots to notify external systems of leads or events.
Widget
A widget is a small self-contained component embedded into a website — the chatbot widget is the visible chat interface visitors interact with, added via a single line of code.
Workflow Automation
Workflow automation uses technology to automate multi-step business processes — a chatbot can capture a lead, create a CRM record, send a follow-up email, and schedule a demo from one conversation.