MySQL

How do I use it?

To use the MySQL Database Connector in your workflow:

Add the MySQL Database Connector

• Insert the node into your pipeline where you want to interact with MySQL databases
• Connect an LLM component to the “llm” invoke connection for query generation

Connect Input

• Connect the input lanes (questions, answers, or table data) to your data sources
• Questions lane accepts natural language queries
• Answers lane accepts JSON data for insertion
• Table lane accepts structured table data

Configure Parameters

• Set MySQL connection details (host, user, password, database, table)
• Adjust connection settings as needed (see tables below)

Connect Output

• The connector outputs transcribed text, formatted tables, or structured answers for further processing

Configuration Parameters

Customize the database connection with the following parameters:

Input Types

The connector accepts three types of inputs for different use cases:

Questions Input

Purpose: Generate and execute SQL queries from natural language

Format: Plain text questions about your data

Example: “Show me all employees with salary greater than $50,000”

Answers Input

Purpose: Insert structured data into MySQL tables

Format: JSON objects or arrays of objects

Example:

[
  {"name": "John Doe", "email": "[email protected]", "department": "Engineering"},
  {"name": "Jane Smith", "email": "[email protected]", "department": "Marketing"}
]

Table Data Input

Purpose: Insert tabular data directly into MySQL

Format: Markdown tables or structured table data

Example: CSV-like data or markdown formatted tables

Output Types

The connector produces three types of outputs for different use cases:

Text Output

• Format: Plain text representation of results
• Use Case: Simple data display or logging
• Example: “Query returned 15 records”

Table Output

• Format: Markdown-formatted tables
• Use Case: Structured data display

Example:

| id | name | email | department |
|----|------|-------|------------|
| 1 | John Doe | [email protected] | Engineering |
| 2 | Jane Smith | [email protected] | Marketing |

Answers Output

• Format: Structured answer objects
• Use Case: Integration with other AI components
• Example: Answer objects containing formatted results

Safety Features

The connector includes comprehensive safety checks to prevent unauthorized database modifications:

Blocked Operations

• DELETE statements
• INSERT statements (when in query mode)
• UPDATE statements
• DROP TABLE/DATABASE
• ALTER TABLE
• TRUNCATE TABLE
• CREATE TABLE/DATABASE
• GRANT/REVOKE permissions
• EXEC/EXECUTE commands
• LOAD DATA operations

Allowed Operations

• SELECT queries (read-only)
• Safe data insertion (when configured for insertion)

Example Use Cases

• Query customer data for analysis and reporting
• Insert processed data from other pipeline components
• Generate reports from database tables using natural language
• Create searchable archives of structured data
• Process and store results from AI analysis
• Extract specific information from large datasets
• Prepare data for machine learning workflows
• Generate real-time analytics from database queries

Best Practices

Model Selection: Use smaller models for quick drafts, larger models for final transcripts

Audio Quality: Higher quality audio produces better transcription results

Chunking: Adjust minimum/maximum seconds based on your content type (shorter for conversational, longer for lectures)

VAD Tuning: Start with balanced VAD and adjust based on your audio environment

For Data Querying

• Be specific with your questions
• Use natural language instead of SQL syntax
• The connector automatically limits to 250 records unless specified otherwise
• Ensure your questions reference actual table and column names

For Data Insertion

• Match your data columns to your table schema
• Use consistent data types for your table structure
• The connector handles batch insertion efficiently
• Validate your data before insertion

For Configuration

• Use strong passwords for database security
• Test connectivity before processing large datasets
• Use connection pooling for optimal performance
• Always backup data before bulk operations

Troubleshooting

Connection Errors

• Verify host, username, and password
• Check network connectivity
• Ensure MySQL server is running
• Verify database exists

Query Errors

• Check if table names exist
• Verify column names in your questions
• Ensure you’re not trying to modify data in query mode

Insertion Errors

• Verify table schema matches your data
• Check data types compatibility
• Ensure required columns are provided

Performance Issues

• Use connection pooling
• Batch large datasets
• Monitor database performance
• Consider indexing for large tables

Dependencies

The MySQL connector requires:

• pymysql==1.1.1 – MySQL Python driver
• cryptography==42.0.2 – Encryption support
• SQLAlchemy==2.0.38 – Database abstraction layer

Integration Notes

• LLM Requirement: An LLM component must be connected for query generation
• Database Compatibility: Works with MySQL 5.7+ and MariaDB 10.2+
• Character Encoding: Supports UTF-8 encoding
• Timezone: Respects database timezone settings

Security Considerations

• All passwords are encrypted in storage
• SQL injection protection through query validation
• Read-only mode prevents accidental data modification
• Connection credentials are never logged
• Automatic session management and cleanup

Performance Tips

• Use connection pooling for high-volume operations
• Batch insert large datasets
• Index frequently queried columns
• Monitor query performance
• Use appropriate data types for your schema

In summary:

The MySQL Database Connector provides flexible, secure database interaction for querying and data insertion operations, with AI-powered natural language query generation, comprehensive safety features, and multiple output formats to fit a wide range of database management and analysis needs.