Plain-English definitions for the most common business intelligence, embedded analytics, and modern data stack terms. Each entry includes a short definition, a concrete example, related concepts, and a link to where the term shows up in real BI products.
This glossary is curated by the Analytify team and updated regularly. If you spot an error or want to suggest a new term, let us know.
Showing all 50 terms.
The discipline of turning operational data into structured insights, dashboards, and reports that guide decisions across every team.
A visual data display consolidating key metrics, charts, and KPIs onto a single screen for at-a-glance monitoring and decision-making.
A quantifiable measurement used to evaluate success against a specific business objective — for example MRR, CAC, or churn rate.
A focused dashboard showing 5-9 key metrics with current value, target, trend, and status — the executive decision tool.
A structured document or screen that aggregates business data into tables, charts, and KPIs to support decision-making at a defined cadence.
Summarising historical data through dashboards and reports to answer "what happened?" — the foundation layer of every BI program.
Representing information through charts, graphs, and maps so people can spot patterns, trends, and outliers faster than reading raw numbers.
BI navigation that lets a user click a summary metric and progressively zoom into more granular detail along a hierarchy.
A centralised database designed to store and analyse large volumes of structured business data, optimised for analytical queries.
A centralised repository storing large volumes of raw data — structured, semi-structured, and unstructured — in its native format on cheap object storage.
A modern architecture combining data lake economics with data warehouse query performance via open table formats like Iceberg and Delta Lake.
An automated sequence of steps that moves data from sources through transformation and into a destination warehouse or analytics tool.
Extract, Transform, Load — a data integration process where raw data is extracted from sources, transformed, and loaded into a destination.
Extract, Load, Transform — a modern integration pattern where raw data is loaded into the warehouse first and transformed inside it with SQL.
Open-source command-line tool that lets analysts and engineers transform data inside a cloud warehouse using version-controlled, tested SQL.
Combining data from multiple sources (ETL, ELT, CDC, streaming, reverse ETL) into a unified view for analytics and operations.
Syncing data from a warehouse to operational tools (CRM, marketing, support) — making the warehouse the operational source of truth.
Ingesting, processing, and querying data within seconds of generation — powering fraud detection, IoT, and live dashboards.
A database workload pattern of fast, small read/write operations supporting transactional applications like e-commerce, banking, and SaaS.
A subject-specific subset of a data warehouse focused on a single department or function (sales, marketing, finance) for faster team-level access.
A centralised set of metric and dimension definitions that sits between your warehouse and BI tool, ensuring every dashboard agrees.
A warehouse design pattern with a central fact table containing measurable events surrounded by dimension tables holding descriptive attributes.
A warehouse design that normalises dimension tables into multiple related sub-dimensions, reducing storage at the cost of more joins.
The central table in a star schema containing measurable business events with foreign keys to dimensions and numerical measures for aggregation.
A table holding descriptive attributes (customer, product, date) that contextualise the events stored in fact tables.
A database workload pattern characterised by analytical queries — aggregations, filters, and slice-and-dice across large historical datasets.
A multi-dimensional structure that pre-aggregates measures across dimensions so analytical queries return in milliseconds instead of minutes.
Stores table data by column instead of by row, dramatically speeding up analytical queries that aggregate or filter on a few columns.
Structured Query Language — the standard language for querying, modifying, and managing data in relational databases.
Data Analysis Expressions — the formula language used in Power BI, Excel Power Pivot, and Analysis Services to define measures and calculated columns.
The data transformation language behind Power Query (Power BI / Excel) for scripting repeatable steps to clean, reshape, and combine data.
Using historical data, statistics, and ML to forecast future outcomes — churn, demand, fraud, lead scoring, predictive maintenance.
Goes beyond predicting what will happen to recommend or execute the best next action via optimisation, simulation, and AI agents.
AI/ML/NLP-powered automation of analytics workflows — discovery, prep, narration, recommendation — including modern GenBI.
Data indexed by time, plus the techniques used to spot trends, seasonality, anomalies, and forecast future values.
Grouping users by a shared starting event (sign-up week, first purchase) and tracking their behaviour over time as a group.
A neural network with billions of parameters trained on massive text data to understand and generate human language.
An AI architecture that augments LLMs with external knowledge retrieval, dramatically reducing hallucinations and enabling source citations.
An autonomous software system using an LLM plus tools, memory, and planning to accomplish multi-step tasks like data investigation.
Generative Business Intelligence — using LLMs and generative AI for natural-language interactions with business data.
Integrating dashboards, charts, and reporting features directly inside a software product so users analyse their data without leaving it.
A software architecture where a single instance serves many customers, with each tenant's data and config logically isolated.
A BI tool that can be fully rebranded with your logo, colours, domain, and product name so end users experience it as native.
API-first analytics architecture where a centralised semantic layer exposes governed metrics for any consumer (BI tools, AI, custom apps).
Database/BI feature that filters which rows a user can see based on identity or role, enforced at query time rather than after.
The framework of policies, ownership, processes, and tooling that ensures data is accurate, secure, well-defined, and used responsibly.
The degree to which a dataset is fit for use — accuracy, completeness, consistency, timeliness, uniqueness, validity.
A centralised, searchable inventory of every dataset, dashboard, and metric — enriched with descriptions, ownership, lineage, and quality.
Creating and maintaining a single, trusted record for shared entities (customers, products, suppliers) across all systems.
Continuous monitoring of data pipeline health, freshness, quality, and lineage to detect issues before they reach dashboards.
Bookmark this page or follow Analytify on GitHub for updates as we expand the glossary.