Getting Started
This section takes you from signing in for the first time to a published, working AI Employee (AIE) — and helps you choose the right kind of AI Employee and the right use case before you build. If you're new to the platform, read these pages in order; if you already know what you want to build, jump straight to Create Your First AI Employee.
If you haven't read it yet, Introduction explains what an AI Employee is and the concepts the platform is built from. This section assumes that foundation and focuses on doing.
In this section
| Page | What it covers |
|---|---|
| AI Employee Types | The interaction types you can build — Chat, Dashboard, and Voice — what each is for, and how the type shapes the AI Employee. |
| First Login | Signing in with SSO, email one-time passcode, or password, and setting up two-factor authentication. |
| Create Your First AI Employee | A start-to-finish walkthrough: create from a template or from scratch, edit the workflow, test it, and publish. |
| Effective Discovery | A framework for scoping a high-impact use case before you build, with playbooks for HR, sales, and support. |
Before you begin
- You have access to an Ema tenant. Your administrator invites you, and your login method (SSO, one-time passcode, or password) is set at the tenant level. See First Login.
- You have a builder role. Creating and editing AI Employees requires a builder role (
builder,builder_admin,system_admin, orenv_admin). Users with theuserrole can run and chat with AI Employees that have been shared with them, but the "Build an AI Employee" button only appears for builders. Roles are covered in Administration. - You have a modern browser. Use a current version of Chrome, Edge, Firefox, or Safari.
The recommended path
- Sign in. Follow First Login to authenticate and, if your tenant requires it, enroll in two-factor authentication.
- Pick a type. Read AI Employee Types to decide whether your use case is a Chat, Dashboard, or Voice AI Employee. The interaction type is set at creation and shapes the rest of the build.
- Build something. Walk through Create Your First AI Employee to create an AI Employee from a template, edit its workflow in the AI Employee builder, test it, and publish.
- Scope the next one well. Use Effective Discovery to choose and validate a use case worth building before you invest in a full deployment.
Let Autopilot do the work. Instead of building by hand, you can describe what you want in plain language and let Autopilot — Ema's in-app assistant — create and edit the AI Employee for you. It's the fastest way to get a first draft.