Generate AI-Powered LinkedIn Posts with Ollama, Image Creation, and Gmail Delivery
Go to WorkflowDescription
This automation flow is designed to generate professional, research-backed LinkedIn posts based on a user-submitted topic and audience, enhance it with a visually aligned image prompt, and then automatically send the post and image via Gmail and/or publish it to LinkedIn.
βοΈ How It Works (Step-by-Step):
π Form Submission (Input Trigger)
A user fills out a form with:
Email
Topic of the Post
Target Audience
This form submission acts as the entry point of the workflow (Form Trigger node).
π€ LinkedIn Post Generation Agent
An AI agent is triggered that:
Uses Tavily API to fetch real-time web search content related to the topic.
Processes the topic and audience data using an Ollama Chat Model.
Generates a well-structured LinkedIn post that includes:
Hook
Educational insight
Professional tone
Source citations
Hashtags and call to action
π¨ Image Prompt Agent
The generated post is passed to a second agent that:
Extracts the core message.
Converts it into a graphic prompt using a different Ollama LLM.
The prompt is written to be used by an image-generation model like DALLΒ·E or Gemma/GPT-based image tools.
πΌοΈ Image Generation
The image prompt is passed to a local image generation endpoint (localhost:8098) or OpenAI (api.openai.com) to generate a visual asset.
The base64 image is converted into a file.
π© Email Dispatch
The post text and the generated image are then sent to the userβs email address via the Gmail node.
The email includes:
The full LinkedIn post
The auto-generated image as an attachment
π LinkedIn Publishing (Optional)
If OAuth is enabled, the post can also be published directly to LinkedIn using the LinkedIn node.
π οΈ Tools Used:
n8n: Orchestration platform
OpenRouter & Ollama: Local/hosted LLMs for post and prompt generation
Tavily API: Real-time web search
OpenAI or Local Endpoint: Image generation from text
SMTP (Gmail): For sending emails
LinkedIn API: For automated posting
π¦ Key Features:
End-to-end automation from form to email/post
Uses real-time research for accuracy
Creative visual generation with professional branding intent
Works locally or with cloud APIs
Modular: Can switch tools or endpoints (OpenAI β Localhost, Ollama β OpenRouter)
π Ideal Use Cases:
Marketing teams needing fast content turnaround
Personal branding assistants
Founders, freelancers, and executives sharing thought leadership
University or organization-wide communication systems