LinkedIn Auto-Post

Maintaining a consistent, high-quality LinkedIn presence for a company demands daily creativity, time and discipline — resources that are always scarce. The goal was to eliminate the overhead of manual content creation while keeping a human in full control of every post published. No fully autonomous bots, no off-brand content: just speed, consistency and quality, powered by AI.

Built a fully automated content pipeline in n8n that fires every morning at 10:00 AM. The workflow selects a topic from a curated list of company-relevant themes, feeds it into a prompt-engineered OpenAI content generator, and delivers a polished draft post directly to the user's Telegram. From Telegram, the user controls the entire editorial flow without leaving the app: • Approve + attach image — sends post and image straight to publishing • Skip today — no post is published that day • Regenerate (same topic) — generates a new draft keeping the original theme • Change topic + regenerate — picks a different topic from the list and rewrites • Edit manually — the user types the final text directly in the chat Once the post is approved, n8n triggers a webhook to Make (formerly Integromat), which handles LinkedIn's API authentication and publishes the post — with the attached image — directly to the company page.

The system bridges three platforms through a clean event-driven architecture: n8n — orchestration layer: cron trigger, topic selection logic, OpenAI API calls with structured prompt engineering, Telegram bot interactions (send, listen, branch on user action), and outbound webhook to Make. Telegram Bot — the human interface: all user interactions happen here. The bot presents the draft, listens for the user's choice, handles image attachments, and routes the decision back to n8n. Make (Integromat) — publishing layer: receives the approved post payload via webhook and handles the OAuth flow and LinkedIn API call to publish to the company page. This separation keeps n8n clean of credential management for external social APIs.

Daily, consistent LinkedIn posts published with minimal human effort — every single one reviewed and approved by a human before going live. Prompt engineering ensures on-brand voice, relevant content and professional tone on every generation. Total time spent per post: under 60 seconds. The system eliminated ad-hoc content scrambling and last-minute posting stress, turning social media management into a lightweight, controlled daily routine. The result is a stronger, more consistent brand presence on LinkedIn with a fraction of the manual effort.