Databricks SQL Agent

Required Parameters

• Language Model: The AI model for processing
• Warehouse ID: Databricks SQL warehouse identifier
• Host URL: Databricks workspace URL
• Access Token: Authentication token
• Schema: Database schema information

Optional Configuration

• Query Options: Query settings
  • maxRows: Maximum rows to return
  • timeout: Query timeout
  • caching: Enable result caching
• Performance: Performance settings
  • concurrency: Concurrent queries
  • retryPolicy: Retry configuration
  • pooling: Connection pooling
• Security: Security options
  • encryption: Data encryption
  • auditLogging: Query logging

Output Format

{
  "result": {
    "query": {
      "sql": string,
      "parameters": object
    },
    "data": {
      "columns": array,
      "rows": array,
      "rowCount": number
    },
    "metadata": {
      "executionTime": string,
      "warehouseId": string,
      "queryId": string
    }
  }
}

Example Usage

const databricksSQLAgent = new DatabricksSQLAgent({
  languageModel: "gpt-4",
  warehouseId: "warehouse_id",
  hostUrl: "https://your-workspace.cloud.databricks.com",
  accessToken: process.env.DATABRICKS_TOKEN,
  schema: {
    database: "sales_db",
    tables: ["orders", "customers", "products"]
  },
  queryOptions: {
    maxRows: 1000,
    timeout: 300,
    caching: true
  },
  performance: {
    concurrency: 5,
    retryPolicy: {
      maxAttempts: 3,
      backoff: "exponential"
    },
    pooling: {
      min: 1,
      max: 10
    }
  },
  security: {
    encryption: true,
    auditLogging: true
  }
});

const result = await databricksSQLAgent.process({
  input: "Show total sales by region for last quarter"
});