AI Assistant

Text2SQL.ai's AI assistant helps you generate SQL queries, fix errors, and optimize your database interactions through natural language conversations.

Starting a New Conversation

There are two ways to start a new conversation with the assistant:

1. Start New Chat

   • Click "New Chat" in the sidebar or navigate to the Home page
   • Enter your question or request in the input field
   • Recommended - Select a database connection for context
   • Press Enter or click the send button to start the conversation

2. Using Templates

   • Choose from pre-built templates for common tasks:
     • Generate SQL from text descriptions
     • Fix SQL syntax errors
     • Optimize SQL queries
     • Generate regex patterns
     • Create spreadsheet formulas
   • Fill in the template details
   • The assistant will process your request using the template format

Ask Follow-ups

The assistant maintains context throughout your conversation, allowing you to:

• Refine your queries
• Ask for explanations
• Request optimizations
• Debug errors
• Get clarification on previous responses

Best Practices

1. Provide Context

   • Configure and link a database connection when possible. Make sure you have selected a type and entered a detailed schema. You can find more information here: Database Connections

2. Clear Communication

   • Be specific in your requests
   • Break complex queries into smaller steps
   • Include example data or expected results when relevant

Example Interactions

1. Generate a Query

   User: Show me all orders from the last 7 days
   Assistant: Based on your schema, here's a query to fetch recent orders:
   
   SELECT *
   FROM orders
   WHERE order_date >= CURRENT_DATE - INTERVAL '7 days'
   ORDER BY order_date DESC;
   

2. Fix an Error

   User: This query isn't working: SELECT * FORM users;
   Assistant: I spotted the error. "FORM" should be "FROM". Here's the corrected query:
   
   SELECT * FROM users;
   

3. Optimize a Query

   User: How can I make this query faster?
   SELECT * FROM orders WHERE EXTRACT(YEAR FROM order_date) = 2024;
   
   Assistant: Here's an optimized version using a more efficient date comparison:
   
   SELECT * FROM orders
   WHERE order_date >= '2024-01-01'
   AND order_date < '2025-01-01';
   

Access Past Conversations

• View your chat history in the sidebar
• Search through previous conversations
• Reuse successful queries from past interactions
• Continue previous conversations where you left off