Input Database

---

🎯 Purpose

The Input Database node connects workflows to your organization’s databases. It retrieves data from tables and makes it available for automation, AI processing, and downstream nodes. Use it to pull customer records, transactions, inventory levels, HR data, or any structured business information.

📤 Outputs

• Retrieved database rows (structured data).
• Metadata about selected database, table, and applied filters.

⚙️ Parameters

Name	Type	Required	Default	Description
Database	Dropdown	Yes	None	Select the database connection.
Table	Dropdown (searchable)	Yes	None	Select the target table within the chosen database.
Column	Data Grid	No	Empty	Configure column purposes: Input (use), Predict (analyze), Ignore (skip).
Filter	String	No	Empty	SQL-style filter condition (e.g., status = 'active').
Limit	Number	No	0	Maximum number of rows to retrieve (0 = no limit).
Sort	String	No	Empty	Sort order (e.g., created_date DESC).
Disable Cache	Boolean	No	false	Disable caching to always fetch fresh data.
Manual Order	Number	No	1	Processing order when multiple Input Database nodes exist.

💡 Example Usage

• 📞 Customer Service: Query the customer_accounts table with filter status='active' to preload customer info for support requests.
• 📈 Sales Leads: Retrieve new leads (lead_date > '2024-01-01'), mark revenue_potential as Predict, and sort by company size.
• 📦 Inventory Management: Pull low-stock items (current_stock < minimum_threshold) and trigger reorder workflows.
• 🏥 Healthcare Records: Fetch upcoming appointments by filtering appointment_date >= TODAY().
• 💳 Fraud Detection: Load recent transactions, marking amount and timestamp columns for Predict analysis.

📘 Best Practices

• Use filters and limits to reduce query size and improve performance.
• Only set relevant columns as Input or Predict to minimize unnecessary data.
• Disable cache for real-time data (e.g., stock levels, transactions).
• Use manual order to sequence multiple Input Database nodes.
• Document filter and sort logic for maintainability.

🧪 Test Cases

• Given: Filter = created_date > '2024-01-01' → Expected: Only recent records retrieved.
• Given: Limit = 50 → Expected: First 50 rows returned in specified sort order.
• Given: Column amount set to Predict → Expected: Value available for downstream forecasting.