Make

Add human review to your Make scenarios.

Make (formerly Integromat) is a visual automation platform. This guide shows how to add human-in-the-loop review to your scenarios using Datashift's REST API and webhooks.

What you'll build

A scenario that generates AI content and routes it through human review before publishing.

Prerequisites

•A Datashift account with an API key
•A Make account (free tier works)
•A review queue created in Datashift

Architecture

The integration uses two Make scenarios:

1. Submit Scenario

Processes data and submits it to Datashift for review.

2. Webhook Scenario

Triggered when the review completes, continues the workflow.

Step 1: Create Webhook Scenario

First, create a scenario that will receive the review result:

Create a new scenario

Click "Create a new scenario" in Make

Add a Webhooks trigger

Search for "Webhooks" and select "Custom webhook"

Create and copy the webhook URL

Click "Add", name it "Datashift Review", and copy the URL

Add a Router module

Route based on approval status to different actions

Step 2: Register Webhook in Datashift

Go to Settings → Webhooks in Datashift:

URLhttps://hook.make.com/xxxxxxxxxx

Eventstask.reviewed

Save the signing secret for webhook verification.

Step 3: Create Submit Scenario

Create a second scenario that submits tasks for review:

Add HTTP Module

Search for "HTTP" and select "Make a request".

URLhttps://api.datashift.io/v1/task

MethodPOST

Body typeRaw (application/json)

Headers

AuthorizationBearer your_api_key

Content-Typeapplication/json

Request Body

json

{
  "queue_key": "content-review",
  "data": {
    "content": "{{1.content}}",
    "platform": "{{1.platform}}",
    "scheduled_time": "{{1.scheduled_time}}"
  },
  "metadata": {
    "make_execution_id": "{{executionId}}",
    "source_module": "{{1.module}}"
  },
  "summary": "Review content for {{1.platform}}"
}

Use Make's variable picker to map fields from previous modules.

Step 4: Configure the Router

In your webhook scenario, add a Router module after the webhook trigger:

Approved Route

Filter condition:

{{data.reviews[0].result}} Contains "approved"

Connect to your publishing module (e.g., post to social media).

Rejected Route

Filter condition:

{{data.reviews[0].result}} Contains "rejected"

Connect to notification module (e.g., send Slack message).

Complete Example

Here's how the complete flow looks:

Submit Scenario

OpenAI: Generate→HTTP: Submit to Datashift

Webhook Scenario

Webhook→Router→Publish/Notify

Accessing Review Data

The webhook payload includes the complete task and review data:

json

{
  "event": "task.reviewed",
  "timestamp": "2026-02-26T10:35:00Z",
  "data": {
    "task": {
      "id": "task_abc123",
      "external_id": null,
      "state": "reviewed",
      "summary": "Review content for Twitter",
      "data": { "content": "...", "platform": "twitter" },
      "metadata": { "make_execution_id": "exec_xyz" },
      "sla_deadline": null,
      "reviewed_at": "2026-02-26T10:35:00Z",
      "created_at": "2026-02-26T10:30:00Z"
    },
    "queue": {
      "key": "content-review",
      "name": "Content Review",
      "review_type": "approval"
    },
    "reviews": [{
      "result": ["approved"],
      "data": {},
      "feedback": "Looks good, ready to publish",
      "reviewer": { "name": "Jane Smith", "type": "human" },
      "created_at": "2026-02-26T10:35:00Z"
    }]
  }
}

Access fields like {{data.reviews[0].feedback}} in subsequent modules.

Tips

Include execution context

Store the Make execution ID in metadata to correlate webhook responses with original runs.

Use Data Stores for state

If you need to preserve data between the submit and webhook scenarios, use Make's Data Stores feature.

Set up error handling

Add error handlers to your HTTP module to catch API failures and retry or alert.

Test with sample data

Use Make's "Run once" feature to test individual modules before activating the full scenario.