EveryRow Make.com Integration

A Make.com app for EveryRow - AI-powered data operations for your workflows.

Features

EveryRow provides intelligent data processing operations:

• Rank - Score and sort rows by AI-evaluated criteria
• Dedupe - Remove duplicate rows using AI matching
• Screen - Filter rows based on complex AI criteria
• Merge - Join two tables using AI-powered matching
• Agent Map - Run web research on each row

Installation

Step 1: Install the EveryRow App

Click the link below to add EveryRow to your Make.com account:

Install EveryRow App

Step 2: Get Your EveryRow API Key

1. Go to everyrow.io and sign up (or log in)
2. Navigate to everyrow.io/settings/api-keys

New accounts get $20 free credit to get started.

Step 3: Create a Connection in Make.com

1. In Make.com, create a new scenario
2. Add any EveryRow module (e.g., "Start Rank Task")
3. Click Create a connection
4. Paste your EveryRow API key
5. Click Save

Quick Start

The easiest way to get started is to import one of our ready-made templates.

Import a Template

1. Go to the templates/ folder in this repo
2. Copy the contents of a template JSON file (e.g., everyrow-rank.json)
3. In Make.com: Create scenario → ⋯ menu → Import Blueprint
4. Paste the JSON and click Save
5. Add your EveryRow connection to each module
6. Click Run once to test

Available Templates

Template	Description
everyrow-rank.json	Score and sort items by criteria
everyrow-screen.json	Filter rows based on AI criteria
everyrow-dedupe.json	Remove duplicates with fuzzy matching
everyrow-merge.json	Join two tables with AI matching
everyrow-agent-map.json	Web research on each row

Usage

Workflow Pattern

EveryRow operations are asynchronous. The recommended pattern is:

[Start Task] → [Get Status] → [Sleep 2min] → [Get Status] → [Get Results]

For longer tasks (like Agent Map with web research), increase the sleep duration.

Example: Rank Products

1. Add EveryRow → Start Rank Task

   • Input Data: Your JSON array of products
   • Task: "Rank by value for money considering price and features"
   • Field Name: "score"

2. Add EveryRow → Get Task Status

   • Task ID: {{previous.taskId}}

3. Add Tools → Sleep (120 seconds)

4. Add EveryRow → Get Task Status again

5. Add EveryRow → Get Task Results

   • Task ID: {{startRankTask.taskId}}

Modules Reference

Start Rank Task

Scores and sorts items based on AI evaluation.

Parameter	Required	Description
Input Data	Yes	JSON array of objects to rank
Task	Yes	Description of ranking criteria
Field Name	Yes	Output field name for the score (default: "score")
Ascending Order	No	Sort lowest to highest (default: false)

Start Screen Task

Filters rows based on complex criteria.

Parameter	Required	Description
Input Data	Yes	JSON array of objects to filter
Task	Yes	Description of filtering criteria

Start Dedupe Task

Removes duplicate rows using AI-powered matching.

Parameter	Required	Description
Input Data	Yes	JSON array of objects to deduplicate
Equivalence Relation	Yes	Natural language description of what makes rows duplicates

Start Merge Task

Joins two tables using AI-powered matching.

Parameter	Required	Description
Left Table	Yes	Primary JSON array of objects
Right Table	Yes	Secondary JSON array to join
Task	Yes	Description of how to match rows
Left Key	No	Optional column name to match on from left table
Right Key	No	Optional column name to match on from right table

Start Agent Map Task

Runs web research on each row.

Parameter	Required	Description
Input Data	Yes	JSON array of objects to research
Task	Yes	Description of research to perform
Effort Level	No	low, medium, or high (default: low)

Get Task Status

Checks the status of a running task.

Parameter	Required	Description
Task ID	Yes	ID returned from Start Task module

Returns: taskId, status (pending, running, completed, failed), artifactId, error

Get Task Results

Retrieves results from a completed task.

Parameter	Required	Description
Task ID	Yes	ID from Start Task module

Returns: Array of result objects (each row as a separate bundle)

Tips

Input Data Format

Input data should be a JSON array of objects:

[
  {"name": "Product A", "price": 100, "rating": 4.5},
  {"name": "Product B", "price": 150, "rating": 4.8}
]

Use {{toString(yourArray)}} to convert data from other Make.com modules to JSON.

Handling Large Datasets

For datasets with many rows:

• Increase the Sleep duration (3-5 minutes)
• Consider using a polling loop with a Repeater module
• Agent Map tasks take longer due to web research

Error Handling

Check the status field from Get Task Status:

• completed - Task finished successfully, get results
• failed - Check the error field for details
• pending / running - Task still processing, wait longer

Support

• EveryRow Documentation
• Make.com Community
• GitHub Issues

License

MIT License - see LICENSE for details.