> ## Documentation Index
> Fetch the complete documentation index at: https://semantiks.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Console Tour

> Walk through the Semantiks Console: how the sidebar is organized, what each area is for, and where to go first.

export const BetaPill = () => <span style={{
  display: 'inline-block',
  marginLeft: '0.35em',
  padding: '0 6px',
  height: '18px',
  lineHeight: '18px',
  fontSize: '10px',
  fontWeight: 600,
  letterSpacing: '0.06em',
  textTransform: 'uppercase',
  borderRadius: '9999px',
  background: '#dbeafe',
  color: '#1d4ed8',
  border: '1px solid #93c5fd',
  verticalAlign: 'middle'
}}>
    Beta
  </span>;

The **Semantiks Console** is built around your AI agent. You switch context with **tabs** at the top of the sidebar—each tab groups a part of the product so you can focus on building the agent, running the contact center, or managing your audience.

## How you move around

* **Pick a tab** (AI Agent, Contact Center, or Audience) to see the menu for that area.
* **Choose an agent** from the selector when you need to work on a specific agent’s configuration or data.
* **Home** (house icon) opens your agent’s overview dashboard—volume, trends, and health at a glance.

***

## AI Agent tab

Everything under this tab is about **designing, connecting, supervising, and improving** your agent.

### Home

Your **starting point** after selecting an agent: key metrics and activity so you know how the agent is performing before you dive into settings.

### AI Agent section

#### Design

Configure *how* the agent behaves and what it can use:

* **Identity** — Name, tone, greeting, audience, availability, and how the agent presents itself.
* **Knowledge** — Documents, FAQs, and sources the agent can rely on in answers.
* **Skills** — Built-in and custom tools (files, FAQs, collect information, human escalation where supported, and your own HTTP tools).
* **Memories** — What the agent can remember about each user across conversations.
* **Workflows** — Structured, multi-step conversation flows. <BetaPill />

#### Playground

A **safe sandbox** to chat with the agent and test changes before they affect live channels.

#### Connect

One place to **go live**: internal channels (e.g. External Chat, Web Bubble), **external integrations** (messaging and tools your org uses), and **API** access for programmatic messaging. Channel-specific setup is also documented under Integration Guides where needed.

#### Supervise

Operational visibility over live traffic:

* **Conversations** — Browse and drill into conversations across channels.
* **Task Log** — See tool and skill invocations step by step (what the agent actually did).

#### Evaluate

Quality and insight on top of raw traffic:

* **Evaluations** — Filtered, scored views for reviewing conversation quality.
* **Hotspots** — Topics and patterns that stand out. <BetaPill />
* **Alerts & Reports** — Thresholds, notifications, and reporting. <BetaPill />

### Proactive AI section

#### Engagement

Reach users beyond reactive chat:

* **Campaigns** — Schedule outbound campaigns (e.g. approved WhatsApp templates and audiences).
* **Follow ups** — Automated re-engagement after silence in a conversation. <BetaPill />

***

## Contact Center tab

Focused on **human agents** working alongside the AI.

### Live Chat

When human escalation is enabled, this is where teams work live conversations:

* **Inbox** — Open, respond to, and take over escalated threads.
* **Performance** — Team metrics (where available). <BetaPill />
* **Escalation Insights** — Why and when conversations move from AI to humans. <BetaPill />

***

## Audience tab

Manage **people** who interact with your agents:

* **Contacts** — Lists and segments you use for outreach and operations.
* **Directory** — Unified view of users who have talked to your agent across channels.

***

## Suggested first path

1. **Home** — Confirm the agent is active and see baseline usage.
2. **Design** — Identity, Knowledge, and Skills so answers are on-brand and accurate.
3. **Connect** — Turn on at least one channel or API path.
4. **Playground** — Validate behavior before going wider.
5. **Supervise** and **Evaluate** — Monitor conversations and improve over time.

For step-by-step tutorials, open the matching section in this documentation from the left navigation.