Glossary

The following is a list of terms used throughout the documentation.

Term	Definition
Agent	An agent is a program or entity that can make decisions, and take actions to achieve specific goals or objectives. The concept of an agent is central to the field of artificial intelligence, particularly in the study of agent-based systems and autonomous agents. An AI agent is a central concept in artificial intelligence, representing an entity with the ability to perceive, reason, and act to achieve specific goals or objectives.
API	API stands for Application Programming Interface. It is a set of rules and protocols that allows different software applications to communicate and interact with each other. APIs define the methods and data formats that applications can use to request and exchange information, enabling them to work together and perform specific tasks seamlessly.
Batch requests	Batch requests refer to a technique where multiple individual HTTP requests are combined into a single request. Instead of sending each request separately, a client can bundle multiple requests into one batch request and send them to the server in a single round-trip. The server then processes each request in the batch and returns a combined response containing the results for each individual request.
LLM	A Large Language Model (LLM) refers to a class of artificial intelligence models that are designed to understand and generate human language. These models are typically based on deep learning techniques, particularly neural networks, and they have a vast number of parameters, making them capable of processing and generating text at an impressive scale.
NER	Named Entity Recognition (NER) is a natural language processing (NLP) technique used to identify and classify named entities within a text into predefined categories. Named entities are specific words or phrases that refer to real-world objects such as people, organizations, locations, dates, quantities, and other named entities that have distinct and well-defined names. The primary goal of Named Entity Recognition is to recognize and extract these named entities from unstructured text, such as articles, social media posts, or research papers, and assign them to appropriate categories. By identifying named entities, NER helps in extracting valuable information and understanding the context of the text, which is particularly useful in information retrieval, text analysis, sentiment analysis, question-answering systems, and more.
Sites Binding	Binding is the process of extracting data from websites which involves an automated methods to gather information from web pages by navigating through the site's structure, locate relevant data, and extract it for further processing or analysis.
Sentiment	In the context of AI, "sentiments" refer to the emotions, opinions, or attitudes expressed in a piece of text or speech. Sentiment analysis, also known as opinion mining, is a natural language processing (NLP) technique used to determine the sentiment conveyed by a given text, such as a review, tweet, customer feedback, or news article.
WYSIWYG	WYSIWYG is an acronym that stands for "What You See Is What You Get." It is a term commonly used in the context of computer software and applications, particularly in web page editors. The content displayed on the screen during the editing process is a close representation of how it will appear in the final output.
Agent Mode	The type of AI assistant configuration that determines its capabilities and behavior. Available modes include Operator (enhanced by documents, internet search, and functions), Programmer (coding assistant with development tools), Assistant (general-purpose with image generation), Orchestrator (plans and executes tasks using other agents), and Desktop (operates on Linux-based systems with browser control).
Agentic Tooling	A feature that allows agents to generate small task lists to accomplish short and medium-term goals autonomously, enabling more sophisticated problem-solving and multi-step execution capabilities.
Code Execution	A capability that enables agents to execute programming code in secure environments, available for Programmer and Operator agents with agentic tooling enabled.
Dynamic Model Routing	An intelligent system that automatically selects the most appropriate AI model for generating responses based on the task requirements and user instructions, optimizing performance and accuracy.
Functions	APIs and database connections that allow agents to consume external data sources and provide suggestions or static responses, automatically enabled for Operator agents.
Knowledge Hub	The centralized repository where documents and information are stored and organized by topics, enabling agents to access relevant knowledge to answer questions and perform tasks.
Orchestrator	A specialized agent mode that can plan and execute complex tasks by leveraging and coordinating other agents, with capabilities for approval workflows and dynamic agent generation.
RAG	Retrieval-Augmented Generation - a technique that provides contextual, knowledge-based responses by retrieving relevant information from the knowledge base before generating answers.
Topics	Organizational categories in the Knowledge Hub that group related documents and information, with agents able to be assigned up to 10 topics to determine their knowledge scope.
Web Widget	A customizable chat interface that can be embedded into web pages, allowing public interaction with agents through configurable appearance settings and themes.