Glossary – Skyone Studio, AI, and Integrations
This glossary consolidates all terms used, maintaining consistent and accurate definitions, with a focus on the context of Skyone Studio, AI agents, language models, integrations, and data.
Technology evolves rapidly, and with it comes terminology that is not always intuitive. Understanding each concept is essential to using the platform efficiently.
This glossary was created to support that understanding, bringing together the key concepts required to navigate and master the Skyone Studio ecosystem.
A
Ad Hoc Reports
Custom reports generated on demand by users to meet specific needs.
Agent
The term Agent is used in two ways within Skyone Studio:
Artificial Intelligence: An intelligent assistant that responds to users and executes tasks based on AI models, context, and skills.
Agent (Terminal): A secure extension of Skyone Studio that can be installed in on-premises environments, utilizing secure communication mechanisms.
Agent Context
A set of information, parameters, and historical data used by the agent to interpret requests and generate responses.
Agent Workflow
A sequence of actions and interactions that define an agent’s behavior.
Agent Workforce
A group of AI agents working collaboratively to achieve complex objectives.
AI (Artificial Intelligence)
The capability of computational systems to imitate intelligent human behavior, such as learning, decision-making, and problem-solving.
AI Agent
A configurable entity that uses language models, context, and skills to interact, process data, and execute tasks autonomously.
Agentic / Autonomous AI
A type of AI agent that operates independently, using generative AI to perform tasks, make decisions, and optimize processes with minimal human intervention.
API (Application Programming Interface)
A set of rules that enables communication between software systems.
API Gateway
A layer that centralizes, manages, and secures access to multiple APIs, acting as a single entry point.
Automation
The use of technology to execute tasks with minimal or no human intervention.
B
Big Data
Large volumes of data characterized by the 5 Vs: Volume, Velocity, Variety, Value, and Veracity.
BI (Business Intelligence)
The process of analyzing data to generate insights that support decision-making.
C
Channel (Publishing Channel)
A platform or medium where an AI agent is made available for user interaction.
Cloud Computing
The delivery of computing services - including servers, storage, databases, networking, software, analytics, and intelligence - over the internet to enable faster innovation, flexible resources, and economies of scale.
Collaboration
Interaction and cooperation between human users and AI agents.
Connectors
Pre-built integrations that connect applications, systems, and data sources.
Consumption
The usage of AI agents and their capabilities.
Context Window
The maximum number of tokens a language model can process in a single interaction.
Conversation History
A complete record of previous interactions between an agent and a user.
Customizable Workflows
AI workflows that can be adapted to meet specific business requirements.
D
Data Collection
The process of gathering and measuring information on variables of interest to answer research questions, test hypotheses, and evaluate outcomes.
Data Democratization
The process of making data accessible across an organization to support data-driven decision-making.
Data Enrichment
The process of enhancing raw data by adding related information or context.
Data Integration
The process of moving, transforming, and harmonizing data across different systems.
Data Lake
A centralized repository that stores large volumes of structured and unstructured data.
Data Mart
A subset of a Data Warehouse focused on a specific business area.
Data Warehouse
A centralized repository of integrated data used for analytics and reporting.
DataOps
A methodology that applies DevOps principles to data management and integration.
Delay
The waiting time before an agent starts responding, especially in voice interactions.
Dimension
A data structure in a Data Warehouse used to categorize and describe facts, such as time, location, or products.
DSS (Decision Support System)
Systems designed to support decision-making processes.
E
ELT (Extract, Load, Transform)
A variation of ETL in which data is transformed after being loaded.
Embedding
A numerical representation of text used for semantic search, comparison, or information retrieval.
ETL (Extract, Transform, Load)
A process for extracting, transforming, and loading data into a Data Warehouse.
EIS (Executive Information System)
Systems designed to provide strategic information to executives and managers.
Execution Data
Data received from connectors or generated during workflow execution.
F
Fact
Quantitative information stored in a Data Warehouse, such as sales or profit figures.
Fine-Tuning
The process of retraining an AI model with specific data to specialize its behavior.
Data Source
A system, database, file, or API from which an agent retrieves information.
G
GenAI (Generative AI)
A type of artificial intelligence capable of generating new content - such as text, images, audio, or code - based on prompts.
Data Governance
A set of practices and processes that ensure data availability, quality, integrity, security, and proper usage in a corporate environment.
Governance and Orchestration
A structure responsible for controlling and coordinating AI agents, regulating autonomy, system access, and compliance with organizational policies.
GPU (Graphics Processing Unit)
A hardware component specialized in parallel processing, widely used to accelerate graphics, machine learning, and AI workloads.
Trigger
An event or condition that initiates a workflow, activating an agent or skill. In integrations, it represents the starting point of execution.
H
High Adaptability
The ability of AI agents to quickly adjust to changes in context or environment.
I
IDP (Intelligent Document Processing)
A technology that uses artificial intelligence to automatically extract, classify, and process information from documents.
Inference
The process of running an AI model to generate responses, classifications, or predictions from input data.
Integration
The connection between AI agents and external systems - such as APIs, applications, or databases -to enable data exchange and automation.
iPaaS (Integration Platform as a Service)
A cloud-based platform used to create, execute, manage, and govern integrations across systems and applications.
K
Kimball (Ralph Kimball)
Creator of a Data Warehouse methodology based on Data Marts and dimensional modeling.
L
Lakehouse
A data management architecture that combines data lake scalability and flexibility with data warehouse structure and governance.
LLM (Large Language Model)
An advanced AI model trained on large volumes of text data to understand, generate, and interpret natural language.
Private LLM
A Large Language Model hosted in a controlled or dedicated environment to ensure enhanced security, data privacy, and regulatory compliance.
LMM (Large Multimodal Model)
An AI model capable of processing and generating multiple data types, such as text, images, and audio, within a single architecture.
M
Master Data Management (MDM)
A strategy used to standardize, centralize, and manage master data within an organization, ensuring consistency and reliability.
Metabase
An open-source business intelligence tool that enables data exploration, analysis, and visualization through dashboards and reports.
Metadata
Data that describes other data, providing context, meaning, and structure.
Model (AI Model / Language Model)
An algorithmic structure trained on data to interpret information, understand natural language, and generate responses, classifications, or predictions.
Multi-channel Publishing
The capability to deploy and make AI agents available across multiple communication channels.
O
ODS (Operational Data Store)
An intermediate repository between operational systems and the Data Warehouse, used to integrate near real-time data.
OLAP (Online Analytical Processing)
A technology designed for complex, multidimensional data analysis to support decision-making.
OLTP (Online Transaction Processing)
Transactional systems optimized for frequent data insertion, updates, and operational processing.
Omnichannel
A multichannel approach that delivers a consistent and integrated customer experience across web, mobile, phone, and physical channels.
Data Organization
A set of practices and structures used to classify, structure, and store data logically, enabling access, analysis, and governance.
P
Parameter
A value provided to configure or execute an operation, process, or model.
Data Pipeline
A set of automated processes responsible for moving and transforming data between systems.
Pipelines
Automated workflows within the iPaaS layer that define sequences of data integration and transformation steps.
Power BI
Microsoft’s business intelligence tool used for data analysis and visualization through reports and dashboards.
Prompt
A text instruction that guides the behavior and responses of an AI agent or model.
Base Prompt
An initial instruction that defines an AI agent’s personality, tone, and default behavior.
Publication
The process of making AI agents and their functionalities available for use in specific channels.
R
RAG (Retrieval-Augmented Generation)
A technique that combines response generation with queries to documents, URLs, or external data sources to improve accuracy and contextual relevance.
RPA (Robotic Process Automation)
A technology that uses software robots to automate repetitive, rule-based business tasks.
Real-Time Processing
The processing of data at the moment it is generated or ingested, enabling immediate responses.
S
Data Segmentation
The process of dividing a dataset into distinct groups based on shared characteristics or criteria.
Silos
Isolated data structures or systems within an organization that hinder information sharing and integration.
Skill
A specific capability or function that an AI agent can perform.
Native Skill
A platform-provided skill that is ready for use and integrated with standard features.
Custom Skill
A skill created or configured by the user to meet a specific use case.
Snowflake Schema
A database schema that normalizes dimensional tables, reducing redundancy but increasing query complexity.
Star Schema
A Data Warehouse schema with a central fact table and denormalized dimension tables, optimized for analytical query performance.
Staging Area
A temporary area in a Data Warehouse where data is loaded and prepared before final processing and storage.
T
Temperature
A parameter that controls the level of randomness and creativity in responses generated by an AI agent or language model.
Token
The basic unit of text processed by language models, representing a word or part of a word.
Tokenization
The process of splitting text into smaller units called tokens so language models can process and interpret it.
Transformation
A general process of modifying data or workflows to achieve a specific objective.
Data Transformation
The process of converting, adjusting, or enriching data from one format or structure to another, forming a core step in data pipelines and ETL processes prior to loading into a Data Warehouse.
U
Intelligent Use of Tools
The ability of AI agents to contextually select and use available tools and skills to interpret information and execute actions accurately.
URL
An address that identifies and provides access to a resource available on the web.
V
Variables
Flow, integration, and context parameters created during configuration. Variables can be passed to operation parameters at runtime and updated via the Update Parameters module. They are classified by scope: Flow, Integration, and Context.
Veracity
A Big Data dimension related to data reliability and trustworthiness.
Volume
A Big Data dimension related to the quantity of data processed.
Last updated
Was this helpful?