Glossary

The words AI and automation tools use, in plain English. 30 terms.

A

Action: An action is a task an automated workflow carries out after it starts, such as sending an email, creating a record, or updating a row in a spreadsheet.
AI agent: An AI agent is software that uses a language model to pursue a goal over several steps, choosing which actions to take and which tools to use.
API: An API (application programming interface) is a set of rules that lets one piece of software request information or actions from another in a predictable way.

Chatbot: A chatbot is software that converses with people through text or voice to answer questions or complete simple tasks, often on a website or in a messaging app.
Context window: A context window is the maximum amount of text, measured in tokens, that an AI model can work with at once, including the conversation, files, and its reply.
Copilot: A copilot is an AI assistant built into an app that helps a person as they work, suggesting drafts, answers, or next steps while the person stays in control.

Embedding: An embedding is a list of numbers that represents the meaning of a piece of text, image, or other content, so software can find items with similar meaning.

Fine-tuning: Fine-tuning is further training of an existing AI model on a smaller set of examples so it handles a particular task or style more consistently.

Guardrails: Guardrails are rules and technical limits placed around an AI system to keep its behavior safe, accurate, and within policy, such as blocking certain topics or requiring approval for actions.

Hallucination: A hallucination is an AI model's confident but false or made-up output, such as an invented fact, quote, statistic, or source that looks believable.
Human in the loop: Human in the loop means designing an AI or automated process so a person reviews, approves, or corrects its work at key points before it takes effect.

Integration: An integration is a ready-made connection between two apps that lets them share data or trigger actions in each other without manual copying.

Large language model (LLM): A large language model (LLM) is an AI system trained on very large amounts of text to predict and generate language, which lets it answer questions, write, summarize, and translate.
Low-code: Low-code describes tools that let people build apps or automations mostly with visual editors, while allowing small amounts of written code for the parts visual tools cannot handle.

Model Context Protocol (MCP): Model Context Protocol (MCP) is an open standard for connecting AI applications to outside tools and data sources, such as files, databases, and business apps, through a common interface.
Multimodal AI: Multimodal AI is AI that can take in or produce more than one kind of content, such as text, images, audio, and video, within the same system.

No-code: No-code describes tools that let people build apps, websites, or automations by using visual editors, drag-and-drop blocks, and settings instead of writing programming code.

OCR (optical character recognition): OCR (optical character recognition) is technology that reads the text in images, scans, and photos and turns it into editable, searchable text.

Prompt: A prompt is the instruction, question, or material given to an AI model to tell it what to produce, such as a request to summarize a document.
Prompt engineering: Prompt engineering is the practice of designing and testing prompts so an AI model produces accurate, consistent, useful results for a particular task.

Retrieval-augmented generation (RAG): Retrieval-augmented generation (RAG) is a method in which an AI system first searches a set of documents for relevant passages, then uses them to write its answer.
RPA (robotic process automation): RPA (robotic process automation) is software that automates repetitive computer tasks by mimicking the clicks and keystrokes a person would make on screen.

Scenario (Make): A scenario is Make's name for an automated workflow, built on a visual canvas by linking modules that each connect to an app or perform a step.

Token: A token is a small chunk of text, often a whole word or part of one, that language models read and write and that providers use to measure usage.
Transcription: Transcription is the conversion of spoken audio, such as a meeting, call, or interview, into written text, which AI tools can do automatically.
Trigger: A trigger is the event that starts an automated workflow, such as a new email arriving, a form being submitted, or a set time of day being reached.

Webhook: A webhook is a way for one app to send information to another automatically the moment something happens, by sending a message to a unique web address.
Workflow: A workflow is a defined sequence of steps that moves a piece of work from start to finish, such as receiving, approving, and paying an invoice.
Workflow automation: Workflow automation is the use of software to carry out the steps of a workflow automatically, so repetitive tasks run without someone doing each one by hand.

Zap (Zapier): A Zap is Zapier's name for an automated workflow, made of a trigger that starts it and one or more actions that Zapier then carries out.