Marketing Analytics
Role: AI Automation Developer

AI LinkedIn Content Engine: Posts & Carousels with Owner Approval

Client: B2B Founder & Executive Brand
3 weeks
7 Technologies

Key Measurable Results

Minutes

Spent by the owner approving posts instead of hours writing & designing

100%

Human-approved before anything goes live on LinkedIn

Post + Deck

Written post and designed carousel slides generated together on schedule

1The Problem

Publishing strong LinkedIn content every day was taking the founder hours—finding a real source worth writing about, drafting the post, and designing a matching carousel. On busy days nothing went out, yet the owner still wanted 100% final approval before anything was published under their name.

2What We Built

Built a 3-stage LinkedIn content engine in n8n that pulls the next topic from Google Sheets, scrapes the real source article via Apify, writes the post and designs carousel slides in Google Drive, and pauses for one-click owner approval before publishing via the LinkedIn API.

The Tools & How We Used Them

n8n & Google Sheets Content Queue

Runs on a schedule, picks the next queued topic from Google Sheets, and logs the final live LinkedIn URL and status back to the sheet after publishing.

Apify Source Scraper

Scrapes the underlying article or source post for each topic so every draft is grounded in real facts rather than generic AI filler.

OpenRouter & Google Drive Carousels

Writes the post copy and slide-by-slide carousel text via OpenRouter, then generates and saves the visual carousel slides into Google Drive.

Owner Approval Gate & LinkedIn API

Holds the finished post and carousel until the owner clicks Approve—then publishes directly to LinkedIn via the official API.

How It Works — Step by Step

1

Structured a Google Sheets topic queue that tracks upcoming ideas, approval status, and published links

2

Integrated Apify to scrape source articles so every post is based on real reference material

3

Configured OpenRouter LLM prompts to draft both the LinkedIn post and structured carousel slide copy

4

Automated carousel slide creation and Google Drive storage for easy review

5

Added a strict human-in-the-loop approval step before publishing via the LinkedIn API

✓The Result

Minutes

Spent by the owner approving posts instead of hours writing & designing

100%

Human-approved before anything goes live on LinkedIn

Post + Deck

Written post and designed carousel slides generated together on schedule

22-Second Walkthrough

See this workflow in action

Watch the 22-second walkthrough showing how this system runs from start to finish.

Skills & Deliverables

n8nLinkedIn APIApifyOpenRouterGoogle DriveGoogle SheetsContent Automation

Project Details

Our Role

AI Automation Developer

Client

B2B Founder & Executive Brand

Industry

Marketing Analytics

Delivery Timeline

3 weeks

Need a similar system built for your team?

Book a free discovery call and we'll map out how to build this for your exact data sources and tools.

More Marketing Analytics Case Studies

View All
Smart Email Campaign Manager: Follow-ups Based on Opens & ClicksVideo

Smart Email Campaign Manager: Follow-ups Based on Opens & Clicks

The client was sending the exact same follow-up email to every lead regardless of what they did with the previous message—someone who clicked a link got the same follow-up as someone who never opened it. Worse, replies were easy to miss, so leads sometimes kept getting automated follow-ups after they had already answered.

View Case Study
AI Music & Sports Content Generator with Email ApprovalVideo

AI Music & Sports Content Generator with Email Approval

The client publishes weekly music and sports content, but keeping up with breaking news, picking trending stories, writing posts, and creating visuals took hours every week—and by the time assets were finished manually, some stories were already stale.

View Case Study
Ecommerce RFM Dashboard: Identified $50k in Churn RisksVideo

Ecommerce RFM Dashboard: Identified $50k in Churn Risks

The business had 693 customers but no clear way to tell them apart. Everyone got the same emails and offers—whether they bought every week or hadn't ordered in almost a year. There was no quick way to see who the best customers were or who had stopped buying.

View Case Study
Social Media Data Stack: Airbyte, Fivetran, BigQuery & dbtVideo

Social Media Data Stack: Airbyte, Fivetran, BigQuery & dbt

The client's analysts were pulling social media data by hand every week—logging into YouTube, Facebook, Instagram, TikTok, and other platforms one by one, exporting CSVs, and pasting numbers into reports. That took 8+ hours every week, and because each platform names and counts metrics differently, reports never matched.

View Case Study
CRM on Autopilot: Daily Lead Pipeline AutomationVideo

CRM on Autopilot: Daily Lead Pipeline Automation

A sales team in Spain was running its entire pipeline by hand. Three people spent their days inside the CRM: checking Gmail to see who had replied, moving leads from one stage to the next, setting follow-up reminders, and fixing contact records. It was slow, error-prone, and the only way to handle more leads was to hire more people.

View Case Study
AI Lead Finder Chatbot (Yelp, Google Maps & US Directories)Video

AI Lead Finder Chatbot (Yelp, Google Maps & US Directories)

Building local business lead lists by hand took hours. To build a single prospect list—like coffee shop owners in Miami—someone had to search Yelp, then Google Maps, then US business directories, and manually open every business's website to hunt for an email address or phone number.

View Case Study
AI LinkedIn Content Engine: Posts & Carousels with Owner Approval