West Side AI

Case Study

LinkedIn Thought Leadership Agent

Auto-writes and publishes expert LinkedIn content with zero broken links

0

Broken links published

140-220

Word count target

5 min

Setup time

PROBLEM

The Challenge

AI-generated LinkedIn posts have broken links, bot-like tone, duplicate hashtags, and no source validation. Most LLM-generated content fails basic fact-checking: URLs expire, shorteners get banned, and the writing reads like a press release. Professionals want consistent thought leadership but can't trust AI output to go straight to publish.

SOLUTION

What We Built

n8n workflow with built-in quality gates. Writes 140-220 word posts with topic and structure rotation. Validates every source URL (HEAD + GET requests, shortener blacklist). Runs an auto-repair loop on broken links — finds alternatives before failing. Humanizes tone with a dedicated pass. Deduplicates hashtags. Won't publish unless all sources pass the validation gate. Schedules via LinkedIn API.

OUTCOME

The Result

Zero broken links in production. Human tone across all published posts. Scheduled publishing on a configurable cadence. 5-minute setup for new topics or personas. The quality gates eliminate the manual review step — it only sends to LinkedIn when it's ready.

Tech Stack

Orchestration
n8n (multi-node workflow)
AI
LLM (OpenAI / Claude, configurable)
Publishing
LinkedIn Person & Organization API
Validation
HTTP HEAD + GET source verification
Scheduling
n8n cron trigger

Want something like this for your business?

Work With Us