AI Glossary | AI Agents, RAG, AI Automation & Analytics Terms

AI glossary: agents, automation and analyticsPlain-language definitions of the terms that come up when an organisation puts AI agents, AI automation and AI-driven analytics into production — written by the ATX Institute team.

Terms

AI agents — AI agents are programs that use a large language model to plan and carry out multi-step tasks — retrieving information, calling tools and taking actions — towards a goal. Unlike a chatbot, which answers one message, an agent decides what to do next based on what each step returned.
Agentic AI — Agentic AI describes AI systems that act with a degree of autonomy — planning steps, using tools and adapting to results — rather than only generating a single response.
AI automation — AI automation is the use of AI models to perform work that previously needed a person — reading, classifying, deciding and drafting — inside an automated process.
AI process automation — AI process automation is automating an end-to-end business process in which AI performs some of the steps, under the same ownership, controls and monitoring as any production process.
AI-driven analytics — AI-driven analytics is the use of AI — mainly large language models — inside the analytics workflow: preparing data, writing queries, classifying free text, building reports and explaining results. You ask in plain language; the tool does the routine work and returns an answer you can check.
Retrieval-augmented generation (RAG) — Retrieval-augmented generation (RAG) is a technique in which relevant passages are retrieved from your own documents and given to a language model together with the question. The model answers from that content, so answers stay current and can cite their sources.
Human-in-the-loop — Human-in-the-loop is a design in which an AI workflow pauses at defined points so a person can approve, correct or reject what the AI proposes before it takes effect.
AI guardrails — AI guardrails are automated checks on an AI model's input or output — scope, factual support, safety and policy — that stop or flag a response before it is used.
Data preparation — Data preparation is the work of collecting, cleaning, joining, reshaping and validating data so it can be analysed or used by AI — typically most of the effort in any analytics project.
PII masking — PII masking replaces or hides personally identifiable information — names, account numbers, national ID numbers, contact details — so data can be analysed, shared or sent to an AI model without exposing the people in it. It ranges from irreversible anonymisation to controlled, reversible pseudonymisation.
Scheduled reports — Scheduled reports are reports that refresh their data, recalculate and deliver themselves — as a file, a PDF or an e-mail — at set times, without anyone running them by hand.
AI business intelligence — AI business intelligence (AI BI) is business intelligence software with AI built into the work itself — writing queries, preparing data, building dashboards and explaining results — rather than a chatbot beside the reports.
KPI dashboard — A KPI dashboard is a single view of the few key performance indicators that show how a business or team is doing, usually with filters to drill into a region, product or period.
PDF report automation — PDF automation is using software to produce, fill or read PDF documents without manual work. In reporting it means PDF report automation: generating recurring reports as PDF files on a schedule — from fresh data, with consistent branding — and delivering them without manual assembly.
SQL reporting — SQL reporting is producing reports directly from saved SQL queries, so every figure can be traced to the query that computed it and re-run with new parameters or on a schedule.
Self-service BI — Self-service BI lets business users build their own queries, dashboards and reports without waiting for a data team, within shared definitions and controls.
AI agent orchestration — AI agent orchestration is coordinating the steps, tools, models and people an AI agent relies on — deciding what runs when, passing data between steps and enforcing the controls around them.
AI governance — AI governance is the set of policies, roles, controls and evidence an organisation uses to make sure its AI systems are approved, monitored, accountable and compliant. Done well, it is built into how AI is developed and run, not kept in a separate policy binder.
EU AI Act — The EU AI Act — Regulation (EU) 2024/1689 — is the European Union's law on artificial intelligence. It sorts AI uses by risk and sets obligations for the providers who build AI systems and the deployers who use them, from outright bans to human oversight, logging and transparency duties.
AI agent observability — AI agent observability is the ability to see what an AI agent did on every run — inputs, retrieved evidence, model calls, outputs, cost and errors — so it can be monitored, debugged and audited.
Sovereign AI — Sovereign AI means an organisation or a country keeps control over where its AI processes data, which models it uses and who can access them. For an enterprise it usually means running AI in its own cloud tenancy, region or data centre, with models reached through its own agreements.

ATX Institute on LinkedIn

This site needs JavaScript for the full experience. Every page above is reachable without it.