Merge

Combines multiple input streams into a single output.

Overview

The Merge node connects to multiple upstream nodes and combines their outputs. Use it to:

• Join data from different sources
• Combine parallel workflow branches
• Append datasets together
• Run SQL queries across multiple inputs

Parameters

Parameter	Description	Required
Number of Inputs	How many input streams to accept	Yes
Mode	How to combine the inputs	Yes

Mode Options

Mode	Description
Passthrough	Output items from one selected input only
Append	Output all items from all inputs
Join	SQL-style join on two inputs
SQL	Custom SQL query across all inputs

Mode: Passthrough

Outputs items from a single selected input, ignoring others.

Parameter	Description
Output Input	Which input to pass through (1-based index)

Use case: Select between workflow branches based on configuration.

Mode: Append

Outputs all items from all inputs in sequence.

No additional parameters required.

Use case: Combine data from parallel processing paths.

Mode: Join

Performs SQL-style joins on exactly two inputs.

Parameter	Description
Left Field	Field from input 1 to match
Right Field	Field from input 2 to match
Join Type	Type of join operation
Conflict Resolution	How to handle duplicate field names

Join Types

Type	Description
Inner	Only matching items from both inputs
Left	All from input 1, matches from input 2
Right	All from input 2, matches from input 1
Full	All items from both inputs
Cross	Cartesian product (all combinations)

Conflict Resolution

When both inputs have fields with the same name:

Option	Behavior
Keep left	Use value from input 1
Keep right	Use value from input 2
Keep both	Prefix with input name
Nested	Nest each input's data

Mode: SQL

Execute a custom SQL query against all inputs using PGlite.

Parameter	Description
Query	SQL query string
Output Mode	How to format results (Rows, Array, Single)
Data Mapping	How to map data (Flatten or JSONB)

SQL Query

Each input is available as a table named input1, input2, etc.:

SELECT * FROM input1
JOIN input2 ON input1.id = input2.orderId
WHERE input1.status = 'active'

Output Mode

Mode	Description
Rows	Each row as separate item
Array	All rows in single array
Single	First row only

Output

Depends on the mode:

• Passthrough: Items from selected input
• Append: All items from all inputs
• Join: Merged items based on join logic
• SQL: Query result rows

Examples

Combine Two Data Sources

Append customers from two tables:

1. Set Number of Inputs: 2
2. Set Mode: Append
3. Connect both Get Table nodes

Join Orders with Customers

[Get Table: Customers] ──┐
                          ├── [Merge: Join] → [Output]
[Get Table: Orders] ─────┘

1. Set Number of Inputs: 2
2. Set Mode: Join
3. Left Field: id (from Customers)
4. Right Field: customerId (from Orders)
5. Join Type: Inner

SQL Query Across Inputs

Calculate order totals per customer:

1. Set Mode: SQL
2. Query:

SELECT
  input1.name,
  SUM(input2.amount) as totalSpent
FROM input1
JOIN input2 ON input1.id = input2.customerId
GROUP BY input1.id, input1.name

Select Active Branch

Pass through only one input based on index:

1. Set Mode: Passthrough
2. Set Output Input: 1 (or 2)

Tips

• Inputs are numbered starting from 1 in the UI
• All inputs must have data before merge executes
• Use Append for simple combining without matching
• Use Join for related data with common keys
• Use SQL for complex transformations and aggregations
• Join mode requires exactly 2 inputs