Artificial intelligence glossary for business.

A type of artificial intelligence that creates new content, such as text, images, audio, or code, from instructions. Businesses use it to draft, summarize, answer questions, or produce material in the brand's voice, always with human review.

An AI system trained on large amounts of text that understands and generates language. It is the engine behind assistants such as ChatGPT, Claude, or Gemini. On its own it does not know your company's data: it has to be connected to it in a controlled way.

A system that receives a goal, decides the steps, and uses tools to reach it: it looks up information, records data, or sends messages. It differs from a chatbot because it acts, not just talks, which is why it works with permissions, limits, and human approval.

Coordinating several specialized agents that split a job, for example research, drafting, and review, under an orchestrator that sequences the steps and hands the result to a person for approval.

A program that talks with customers or staff by text or voice. AI assistants understand open questions and answer with the company's information; when they cannot resolve something, they should hand the case to a person.

Using software so the repetitive tasks in a process move forward without manual work, such as sorting requests, copying data between systems, or sending reminders. With AI, tasks that require reading or interpreting text can also be automated.

A technique that lets a language model answer with your company's information: it first searches your documents for relevant passages and then writes the answer from them, citing the source. It reduces made-up answers and avoids retraining the model.

An answer from an AI model that sounds convincing but is false or unsupported. It is reduced with good sources, precise instructions, limits on actions, and human review of important decisions.

An instruction given to an AI model to get a result. In a business system, prompts are designed, tested, and versioned like any other piece of software.

A design principle in which a person reviews or approves the important actions of an AI system before they take effect, such as sending a message, publishing content, or recording a payment.

Using AI to extract data from invoices, orders, contracts, or forms, validate it, and record it in your systems. Documents that break the rules go to a person for review.

An API is the door a system offers so another program can read or write information. Integrating tools through APIs lets an agent query your CRM, ERP, or calendar without copying data by hand.

An open standard that lets AI models connect to tools and data sources in an orderly way, with defined permissions. It makes it easier for one agent to work with several of your company's systems.

A measurement of how a process performs before changing it: time, errors, quality, or cost. Without a baseline you cannot show whether an AI solution improved anything.

A limited version of a solution used to test an idea with real cases before investing in the full system. It is not a production-ready product; it is a tool for deciding.

The ability to see what a running system does: what it answered, how long it took, what it cost, and where it failed. In AI systems it is the basis for fixing errors and controlling spend.

Search engine optimization. Technical and content practices that help a site appear in Google and Bing results when someone searches for what a company offers.

Answer engine optimization. It prepares content so assistants and search engines use it as a direct answer: clear questions, short answers up front, and structured data.

Generative engine optimization. It aims for models such as ChatGPT, Gemini, Perplexity, or Google's AI answers to understand, cite, and recommend a company. It depends on citable content, a consistent brand entity, and open access for AI crawlers.

A standard-format description of what a page contains: a company, a service, a frequently asked question, or an article. It helps search engines and models read the site without ambiguity.

A text file at the root of a site that summarizes for AI models what the company is, what it offers, and where each important page lives. It complements the site; it does not guarantee showing up in answers.

AI models are billed by the amount of text they process, measured in tokens. A system's monthly cost depends on usage volume, so it is controlled with caps and tracking.