---
title: "x101 Glossary: Enterprise AI Terms Explained"
url: https://www.x101.ai/glossary
description: "Plain-language definitions of the 34 enterprise AI terms x101 uses, from agentic orchestration and RAG to governance, MCP and voice agents."
---

# Enterprise AI terms, in plain language.

The vocabulary used across this site and inside the platform, defined without marketing gloss. 34 terms.

Last updated: 22 September 2026

## All terms

### A

**A2A** (Standard): An open standard that lets AI agents discover, talk to and delegate work to each other across vendors. x101 supports it natively.

**ABAC** (Security): Access control that decides what a user can see or do based on attributes such as role, department and data sensitivity, not just a fixed role.

**Agent** (AI): An autonomous AI worker that can reason, use tools and take actions toward a goal, rather than only answering questions.

**Agent Studio** (x101): x101's no-code visual designer for building, configuring and shipping agents, with persona, tools, memory and guardrails.

**Agentic Orchestration** (x101): Coordinating a team of specialist agents in parallel across systems to complete one task, then synthesising a single answer.

**Agentic Workflow** (x101): A workflow where each step can reason and decide for itself, not just follow a fixed rule.

**Air-gapped** (Deploy): A deployment fully isolated from external networks for maximum security. x101 can run air-gapped.

**Audit Trail** (Governance): A complete, signed log of each action, cost and token, per user and per agent, for compliance and replay.

### C

**Citation & Grounding** (AI): Tying an answer to its exact source (document, page, version) so it can be trusted and verified.

**Command Centre** (x101): x101's single conversational screen to ask, decide and act across connected systems.

**Connector** (Integration): A pre-built integration that lets x101 read from and act inside an external system such as a CRM, ERP or ticketing tool.

**Context Window** (AI): The amount of text a language model can consider at once. Large jobs can exceed it, which is why quality drops at scale without orchestration.

### D

**Deep Research** (x101): Automated multi-step research that scans many sources in parallel and returns a cited, formatted report in minutes.

### E

**Embedding** (AI): A numeric representation of text that lets a system find content by meaning. It powers semantic search and retrieval.

### G

**Guardrails** (Governance): Automated checks that keep agent inputs and outputs within policy, covering safety, PII and scope.

### H

**Hallucination** (AI): When an AI produces plausible but unsupported content. Grounding and citations reduce it.

**Human-in-the-loop** (Workflow): A workflow design where a person approves or steps in at defined points before an action completes.

### I

**Inference** (AI): Running a trained model to produce an output. It is the runtime cost of using a language model.

### K

**Knowledge Retrieval** (x101): Finding the exact relevant passage across enterprise documents, grounded in the source and cited. See RAG.

### L

**LLM** (AI): Large Language Model: the model that understands and generates language. x101 is model-agnostic and runs any LLM.

### M

**MCP** (Standard): Model Context Protocol: an open standard for connecting AI models to tools and data sources. x101 supports it natively.

**MTTR** (Ops): Mean Time to Resolution: the average time taken to resolve an incident. A core IT and security operations metric.

**Multi-agent** (AI): Several specialised agents working together under an orchestrator, each handling part of a task.

### O

**On-premises** (Deploy): Running software inside an organisation's own data centre rather than a public cloud. x101 supports it.

**Orchestrator** (x101): The component that decomposes a task, dispatches agents, and synthesises their results into one answer.

**OSS / BSS** (Telecom): Operations and Business Support Systems that run telecom networks and the commercial side of a carrier.

### P

**PII Masking** (Governance): Automatically hiding personally identifiable information in inputs and outputs to protect privacy.

### R

**RAG** (AI): Retrieval-Augmented Generation: grounding a model's answer in retrieved source documents rather than its memory alone, so answers stay accurate and citable.

**RBAC** (Security): Role-Based Access Control: granting access based on a user's assigned role.

**RCA** (Ops): Root Cause Analysis: identifying the underlying cause of an incident rather than just treating the symptom.

### S

**Semantic Layer** (Data): A unified, business-friendly model of data so that a question maps to the right meaning across systems.

**SIEM** (Security): Security Information and Event Management: a system that aggregates security logs and alerts for monitoring and investigation.

### V

**Vector Search** (AI): Finding content by meaning using embeddings, rather than exact keyword matching.

**Voice Agent** (x101): An AI agent that operates over a phone call, with live data access and the ability to take action, hands-free.

## Still unsure what something means?

A 20-minute conversation with a solutions engineer clears up more than a glossary can. Bring the terms that matter to your evaluation.

- [Book a demo](https://www.x101.ai/contact)
- [Explore the Platform](https://www.x101.ai/platform)