CodeVerse Soban: Teaching the Next Generation of Developers

Long before CodeFusion AI was a platform, I was sharing my learning journey on YouTube via CodeVerse Soban.

My mission has always been simple: "Your ultimate destination to learn coding smartly in Urdu/English." I believe that high-tech concepts shouldn't be gated behind language barriers or complex jargon.

From Docker to AI: Breaking Down the Stack

On the channel, I've covered everything from:

Minikube and Helm deployments for cloud-native apps.
Dockerizing Next.js for production environments.
Building AI Chatbots using the same agentic principles that power CodeFusion.

Bridging the Gap

Why do I continue to teach while building a SaaS? Because a tool is only as good as the person using it. By teaching developers how to use FastAPI, Neon DB, and Clerk, I am preparing them to get the most out of CodeFusion AI.

CodeFusion AI helps you build the structure, but CodeVerse Soban teaches you the why behind every line of code.

Join the Community

With over 1,100 subscribers and growing, we are building a community of "smart coders." Whether it's deploying backends on Hugging Face or mastering the latest Next.js features, we do it together.

Check out our latest tutorials and join the mission to make high-end engineering accessible to everyone.

📺 New Video: Personal AI Employee Part 4

Facebook & LinkedIn post automation prompt given below:

🤖 AI Social Media Automation: Master Implementation Prompt

Give the following prompt to an AI assistant (like Claude or GPT) to replicate this exact system.

📝 The Prompt

"I want to build a Semi-Autonomous AI Social Media Manager that uses a 'Human-in-the-Loop' (HITL) architecture. The system should be built using Python and Playwright. Here is the core specifications:

1. Unified Folder Architecture

Create a folder structure to manage task flow:

/Pending_Approval: AI saves drafted posts here (Markdown with YAML meta).
/Approved: Human moves files here to give permission to post.
/Done: AI moves files here after successful posting.
/session: Stores browser session data for persistent logins (Login once, save forever).
/Logs: For error screenshots and event logging.

2. The Muscle (Social Media Executor)

Implement social_media_executor_v2.py:

Use Playwright with launch_persistent_context directed at the /session folder.
LinkedIn Logic: Handle 'Start a post' button detection and editor filling.
Facebook Logic: Handle multi-step posting (Next -> Post/Share). Use keyboard.type for reliability.
Robustness: Take screenshots to /Logs on any failure.

3. The Brain (Master Orchestrator)

Implement master_orchestrator.py:

Continuously monitor /Approved folder.
When it sees a POST_... file, it triggers the Executor.
If successful, move to /Done. If fail, retry up to 3 times then cooldown.

4. Trigger Scripts (Draft Generation)

Implement trigger_posts.py scripts that:

Generate a Markdown post with metadata (platform, content).
Save it directly into /Pending_Approval.

5. Full Workflow (The 24/7 Loop)

Login Session: Run the script once manually to log in to LinkedIn/FB. The session is saved in /session.
Perception/Generation: Run a trigger script to generate a draft in /Pending_Approval.
HITL Approval: Manually review and move the file to /Approved.
Execution: The Master Orchestrator detects the file and performs the actual browser automation to post it."

🚀 How to use this

Copy the text above and give it to your AI assistant.
Ask it to "Generate the Python code for this entire system."
It will create the watchers, triggers, and executors we built!"

🎥 Watch & Learn

CodeVerse Soban: Why I Teach Coding Smartly in Urdu & English