AI Chat Sessions & Prompts

{productName} includes a powerful AI Chat Engine that allows you to interact with your data using natural language. Every conversation is persisted, allowing you to resume work across sessions, share insights, and build a library of reusable prompt templates.

[try:/aiChats|Open AI Chats|bot]

Overview

The AI Chat system is built on three core components:

AI Chats: Persistent containers for your conversations.
AI Messages: Individual exchanges within a chat session.
AI Prompts: Reusable templates for common tasks like data cleaning, translation, or SEO optimization.

Conversations are context-aware. When you start a chat from a specific product or grid, the AI automatically understands the data you are looking at.

Viewing Past AI Chats

You can access your full conversation history at any time to review previous suggestions or continue a deep-dive analysis.

1. Open the **AI Features** section in the sidebar.
2. Click on **AI Chats** [navigate:/aiChats|AI Chats|bot].
3. Use the search bar or filters to find a specific session by title or date.
4. Click on a chat row to open the full conversation history.
5. To rename a session for better organization, edit the **Title** field directly in the chat header.

Resuming Conversations

Unlike standard AI assistants, {productName} chats are never “lost” when you close your browser.

Persistent History: Every message, including AI-generated code snippets and data previews, is saved.
Context Preservation: If a chat was started with a specific filter or selection, that context is restored when you resume.
Actionable History: You can still click buttons generated by the AI in previous sessions (e.g., “Apply Filter”) to interact with your live data.

Prompt Templates

Standardize your AI workflows by saving successful instructions as AI Prompts. This ensures consistency across your team and saves time on repetitive tasks.

Creating a Template

1. Navigate to **AI Prompts** [navigate:/aiPrompts|AI Prompts|workflow].
2. Click **Add Prompt** [badge:plus|success].
3. Define the **Template Type**:
    - `massUpdate`: For bulk editing records.
    - `importFile`: For intelligent data mapping during imports.
    - `aiMassUpdate`: For complex, multi-field AI transformations.
4. Enter your prompt text using placeholders if necessary.
5. Save the prompt to make it available in the AI Chat and Mass Update interfaces.

Using Templates

When starting a new chat or performing an AI Mass Update, click the Templates icon to pick from your saved library.

Understanding Chat Types

The AI Chat adapts its behavior based on the task at hand. You may encounter these specialized chat types:

Chat Type	Purpose
`importer`	Assists with complex file imports and column mapping.
`filter`	Helps you build precise `WHERE` expressions using natural language.
`mapping`	Guides the configuration of data mappers between systems.
`script`	Generates and explains automation scripts.
`exporter`	Helps define custom export formats and logic.
`ql`	Direct interaction with the {productName} Query Language.
`batch`	Handles large-scale operations across thousands of records.

Tips for Better Results

**Be Specific:** Instead of "Fix these products," try "Update the description for all products where the brand is 'Nike' to include 'Official Store' at the end."

**Context Matters:** If you want the AI to look at specific data, make sure you have those records selected or filtered in the grid before opening the chat.

**Verify AI Suggestions:** While the AI is highly capable, always review generated filters or mass update previews before applying them to your production data.

Next Steps

Now that you’ve mastered AI chats, learn how to use natural language to find exactly what you need in the product grid.

[navigate:/help/en/user-flows/ai-filtering|Continue to AI Filtering|filter]

[navigate:/help/en/user-flows/ai-mass-update|Smart Bulk Updates with AI|zap]
[navigate:/help/en/user-flows/ai-prompt-enhancement|Enhance your AI prompts|wand]
[navigate:/help/en/ai-credit|What Is AI Credit?|sparkles]