> Source: https://builder.ema.ai/v2/introduction
> Title: Introduction

# Introduction

Ema is a platform for building **AI Employees** — configurable automations that read your enterprise data, decide what to do, and carry out multi-step work end to end. This section gives you the foundation you need before you build: what an AI Employee is, the concepts the platform is made of, where everything lives in the interface, and a shared vocabulary for the rest of the documentation.

You don't need any prior Ema experience to read this section. It's written for builders, administrators, and anyone evaluating the platform for the first time.

## In this section

Page

What it covers

[What Is Ema?](/builder/v2/introduction/what-is-ema)

The platform overview — AI Employees, how they run, the interfaces they expose, and EmaFusion™, the model layer underneath.

[Key Concepts](/builder/v2/introduction/key-concepts)

The building blocks: AI Employees, workflows (a DAG of typed nodes), agents, knowledge bases, integrations and Tools, human-in-the-loop, and roles.

[Platform Tour](/builder/v2/introduction/platform-tour)

A walkthrough of the actual interface — the sidebar, the AI Employees workspace, the AI Employee builder, Integrations, and the Admin area.

[Glossary](/builder/v2/introduction/glossary)

An alphabetical reference for every platform term used across the documentation.

## How the platform fits together

At the center of Ema is the **AI Employee** (AIE) — a unit of automation you build, test, publish, and run. Every AI Employee is defined by a **workflow**: a directed acyclic graph (DAG) of typed nodes that the workflow engine validates and executes. The `agent` nodes run **agents** (the reasoning steps) and can pause for **human-in-the-loop** approval; the edges between nodes branch the path on conditions; and `publish` nodes emit the results. Agents draw on **knowledge bases** for retrieval and on **integrations** for **Tools** that read from and write to external systems. Underneath every LLM call sits **EmaFusion™**, Ema's model layer, which selects an appropriate model per request.

You build all of this in the **AI Employee builder**, and you can have **Autopilot** — Ema's in-app assistant — build and edit it for you through a conversation.

## What's next

-   Start with [What Is Ema?](/builder/v2/introduction/what-is-ema) for the big picture, then read [Key Concepts](/builder/v2/introduction/key-concepts) to learn the vocabulary.
-   When you're ready to build, continue to [Getting Started](/builder/v2/getting-started) and [Core Concepts](/builder/v2/core-concepts).
-   For the in-app assistant that can build for you, see [Autopilot](/builder/v2/autopilot).