---
title: "Using the Validatar AI Chat Interface"
slug: "using-the-validatar-ai-chat-interface"
updated: 2026-01-16T09:49:33Z
published: 2026-01-16T09:49:33Z
canonical: "docs.validatar.com/using-the-validatar-ai-chat-interface"
---

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

# Using the Validatar AI Chat Interface

The Validatar AI Chat Interface allows you to interact with your data environment using natural language. You can ask questions, generate documentation, and receive recommendations directly within the Validatar UI—without leaving your current workflow.

This guide walks through how to access the AI Chat, start conversations, and use inline AI actions across the platform.

---

## Prerequisites

Before using Validatar AI, ensure that:

- Your Validatar instance includes the **AI feature** and it is enabled.
- Your user account is licensed and within its **monthly token usage limits**.
- You are working in a **cloud-based Validatar instance** (AI is not currently supported for on-prem/server instances).

If AI is not available, contact your Validatar administrator.

---

## Opening the AI Chat Panel

After logging in to Validatar:

1. Locate the **AI Chat icon** in the upper-right corner of the application header.
2. Click the icon to open the AI Chat panel.

The chat panel opens as a resizable side panel and can be closed at any time using the **X** icon in the top-right corner.

---

## Chat Panel Views

The AI Chat panel includes three primary views:

### Home View

The Home view appears when you open the chat panel or start a new conversation.

From here, you can:

- Click a **predefined prompt** to quickly start a conversation.
- Enter your own free-form prompt to begin a new chat.
- Navigate to **All Chats** to view your conversation history.

---

### Chat View

The Chat view is where active conversations take place.

Key capabilities include:

- **Submitting prompts** Type your question or instruction in the input box and click **Send**.
- **Automatic context awareness** The AI automatically receives context about:
  - Your current page
  - The active data object (data source, schema, table, or column)
- **Chat subject editing** Hover over the chat subject in the header and click the pencil icon to rename it.
- **Feedback controls** Use the thumbs-up or thumbs-down icons on AI responses to provide feedback.
- **Optional diagnostics** If enabled by your admin, token usage details may appear below responses.

---

### Chat History View

The Chat History view shows your past conversations.

From this view, you can:

- Click a conversation to reopen it.
- Archive a conversation using the trash icon.

Archived chats are removed from the active list but retained for audit and analysis purposes.

---

## Using Inline AI Actions

In addition to the chat panel, Validatar AI is integrated directly into many areas of the UI through **Inline Actions**.

These actions automatically open the chat panel and pre-populate a prompt based on the current object.

### Common Inline AI Locations

Inline AI actions are available in the following locations:

- **Data Source Metadata** Catalog → Data Source → Info tab → **Auto-Document**
- **Schema Metadata** Catalog → Data Source → Schema → Info tab → **Auto-Document**
- **Table Metadata** Catalog → Data Source → Schema → Table → Info tab → **Auto-Document**
- **Column Metadata** Catalog → Data Source → Schema → Table → Column → Info tab → **Auto-Document**
- **Test Recommendations**
  - Table-level: Table → Tests tab
  - Column-level: Column → Tests tab

If the chat panel is already open, using an inline action will automatically start a **new conversation** with the relevant context applied.

---

## How the AI Uses Context

Each time you submit a prompt, Validatar AI automatically includes:

- Your selected **AI agent and model**
- The **current page and object context**
- A configurable window of **recent chat history**

This allows the AI to provide more accurate, relevant, and actionable responses without requiring you to restate details manually.

---

## Best Practices

- Be specific in your prompts (e.g., “Generate documentation for this table” vs. “Document this”).
- Use inline actions whenever possible to ensure the AI has full object context.
- Rename chat subjects for easier navigation later.
- Provide feedback on responses to help improve AI quality over time.

---

## Troubleshooting

If you are unable to submit prompts or see AI features:

- Confirm your instance is licensed for AI.
- Verify that your monthly token limit has not been exceeded.
- Refresh the page to reload AI configuration.
- Contact your Validatar administrator if issues persist.