Create LinkedIn Content from Telegram using GPT-4 and Dumpling AI

This automation helps you turn quick ideas into polished LinkedIn posts without manual effort. You simply send a text or voice message to your Telegram bot, and the system will research, generate context, design a visual, and save everything neatly in Airtable. You also receive a confirmation back in Telegram.

---

Benefits of this Automation

• Save time: No more spending hours researching and formatting content ideas.
  
• Consistency: You always get a professional, LinkedIn-ready post with structured text and a matching image.
  
• Simplicity: Just drop your idea into Telegram—everything else runs automatically.
  
• Scalability: You can create multiple posts per week without extra effort.
  
• Collaboration: All outputs are stored in Airtable for easy review and approval by your team.
  

Step-by-Step Setup

Step 1: Trigger from Telegram

The workflow starts with the Telegram Trigger node. Whenever you send a message (text or voice) to your Telegram bot, it activates the workflow.

• If you type a message, it captures the text.
  
• If you send an audio note, it downloads the file for transcription.
  

Benefit: This makes it easy to feed ideas from anywhere without opening n8n or Airtable.

Step 2: Detect Message Type

A Switch node checks whether the input is text or audio.

• If text, it goes directly to processing.
  
• If audio, the file is downloaded and passed to OpenAI Whisper for transcription.
  

Benefit: Flexible input—use whichever is faster for you at the moment.

Step 3: Transcribe Audio to Text

For audio messages, the Transcribe Audio to Text node uses OpenAI’s Whisper model to convert speech into clean text.

Benefit: Capture spontaneous spoken ideas, even when you don’t have time to type.

Step 4: Generate Ideas, Context, and Image Prompt

The LangChain Agent node with GPT-4 takes your input text and runs the following process:

1. Researches the topic with the Dumpling AI Web Search tool (limited to 3 sources).
   
2. Refines the raw input into a clear, short content idea.
   
3. Creates 2–4 sentences of supporting context.
   
4. Crafts a descriptive image prompt.
   

Benefit: You get a LinkedIn-ready content block that is professional, concise, and relevant.

Step 5: Clean the Image Prompt

The Code Tool node removes unnecessary line breaks or formatting from the image prompt so that it works smoothly with the image generation tool.

Benefit: Ensures that the generated image looks clean and professional.

Step 6: Generate an Image

The Image Generation Tool node calls Dumpling AI with the cleaned prompt and generates a LinkedIn-appropriate visual (1:1 ratio, high quality).

Benefit: You always have a branded visual to go with your post, avoiding stock-image fatigue.

Step 7: Save Results to Airtable

The output is structured as JSON in the format:

{

  “ideas”: “string”,

  “context”: “string”,

  “image”: “URL”

}

This data is stored automatically in Airtable with columns for Ideas, Context, and Image URL.

Benefit: Keeps your content organized and easy to review, edit, or schedule.

Step 8: Send Confirmation Back to Telegram

Finally, a confirmation message is sent to Telegram letting you know:

• Your content has been created.
  
• The idea, context, and image are saved in Airtable.
  
• You can review and approve them anytime.

Benefit: Immediate feedback without needing to log into Airtable right away.

Final Result

Each time you share an idea on Telegram, you instantly get:

• A short refined content idea.
  
• 2–4 sentences of professional context.
  
• A matching AI-generated LinkedIn-ready image.
  
• Everything saved in Airtable, with confirmation in Telegram.
  

This makes it incredibly easy to maintain a steady LinkedIn presence with minimal effort.

Download the blueprint used in this blog post

Click here to access the blueprint.