Matrices VideoMatrix AccountingMatrix WorkMatrix PeopleMatrix LegalMatrix SalesMatrix eCommerceMatrix Roadmap FinanceMatrix MarketingMatrix CustomerSuccessMatrix HealthMatrix OpenClawMatrix Resources What is agentic AI AI workflow automation Enterprise AI platform Event-driven orchestration Digital twin AI AI agent comparison What is supervised AI AI agents vs chatbots Human-in-the-loop AI AI operations layer Choosing your first AI workflow Glossary Pricing About Studio ↗ Contact

The AMatrix glossary.

The vocabulary of AMatrix and the AI matrix model, defined plainly. If a term on the site is unfamiliar, it is explained here.

AI agent
A software component that pursues a goal autonomously — perceiving context, choosing an action, executing it, and observing the result, repeating until the goal is met.
Agentic AI
AI that acts toward a goal rather than only responding to a prompt. It plans, takes actions, and adapts, deciding what to do next.
Matrix
An AMatrix product: a complete, versioned software product for one business domain, run by a roster of supervised AI agents.
AMatrix
AMatrix Inc., the Montreal company that builds and operates AI matrices — and the platform those matrices run on.
Ethics rubric
The supervision model applied to every matrix: the axes an agent is scored on, the practices it is forbidden, and the gates it must pass before a consequential action.
Orchestration
Coordinating many agents and steps into a dependable workflow — routing work, running steps in parallel, retrying failures, and enforcing gates.
Event-driven
A design where workflows start in response to real events — a new record, a message, a completed task — rather than on a fixed schedule.
LLM
A large language model — the general-purpose AI model that provides the reasoning capability an agent uses.
LLM-agnostic
Not tied to one model vendor. An AMatrix matrix runs on SovrinOS or any connected LLM, and can switch without re-engineering.
SovrinOS
AMatrix sovereign GPU inference, run in Montreal. The optional sovereign choice for matrix inference when full data residency is required.
Sovereign inference
Running AI inference on infrastructure you control, rather than sending data to a third-party model provider.
Pipeline
A matrix structure where work flows through ordered stages — for example VideoMatrix: script, storyboard, inference, render, approval, publish.
Agent roster
A matrix structure organized as a set of named agents, often grouped — for example AccountingMatrix's general-ledger, bank, and crypto groups.
Module
A unit within a modular matrix — a self-contained area of capability, such as a SalesMatrix module for outreach or for pipeline management.
Pillar
A major division of a pillared matrix — for example LegalMatrix's intellectual-property, corporate, and family-law pillars.
Supervision gate
A required checkpoint in a workflow where a consequential action must pass a check, or a human approval, before it proceeds.
Conversational agent
An agent you interact with directly in natural language, such as VideoMatrix's production assistant and cost analyst.
Ready / Coming soon
The two public tiers of the matrix catalog: ready matrices are available now; coming-soon matrices are in active development.
amatrix.studio
AMatrix's sister studio for fixed-price custom builds beyond the matrix platform.

See it in production

AMatrix builds these ideas into real software — twelve AI matrices for real business domains.