End-to-end AI system for generating, designing, and scheduling social media content automatically

Easy Navigation

The Challenge
Skills
The Solution
The Result

The Challenge

Guru Labs needed a scalable way to produce and publish high-quality social media content without relying on manual design, writing, or editing. Content creation was time-consuming, inconsistent, and difficult to scale across their 5 internal brands.

They needed a system that could:

• Reduce manual workload
• Generate content aligned with each brand’s voice and style
• Produce multiple content formats (static posts, carousels, videos)
• Automate image design, video scripting, voice overs, video content, captions, and hashtags
• Perform quality checks
• Schedule everything automatically

How I Work

A blend of AI, automation, and systems design that turns content creation into a scalable, predictable engine.

Step 1: Brand Voice & AI Content System
I prompted the AI element to understand the brand’s voice and style and created prompts based off the brand voice for different types of content (images, carousels, captions, video scripts, hashtags) and set up quality checks so everything stayed consistent.

Step 2: Automation & Workflow Architecture:
I built multi-step automated systems using Make.com that connect multiple APIs into a single workflow, from content generation to validation, formatting, and scheduling.
The system:

• Generates content from AI prompts
• Outputs captions and scripts to Google Sheets for real-time visibility and review
• Runs automated quality checks with rejection and regeneration logic
• Processes content into multiple formats (text, images, video)
• Connects directly to a scheduling platform to queue and publish content automatically

This removed manual handoffs and allowed content to move from idea to approved to scheduled without human intervention.

Tools & Technologies
(The ecosystem I used to achieve this)

• Make.com (multi-step content workflows, automation logic)
• OpenAI / ChatGPT (prompt frameworks, caption + script + hashtag generation, quality control)
• Creatify & Kling API (AI video generation)
• Placid API (Carousel generation)
• Metricool (Automated scheduling + publishing)
• Ideogram API (Image generation)
• Google sheets (For real-time visibility and review)
• Google Workspace (Brand documents, content storage)

The Solution

Build a Complete AI Content Engine From Scratch

1. Created Brand Identity Documents

• Compiled tone, style, messaging, audience, and visual guidelines
• Used brand docs as the “source of truth” for AI generation
• Ensured all content matched each brand’s personality

2. Engineered OpenAI Prompt Frameworks Based on Brand Guidelines
Developed structured prompts for:

• Social media captions
• AI Image prompts
• Static posts
• Carousel posts
• Short-form video scripts
• Hashtags

This ensured every piece of content was on-brand, strategic, and consistent

3. Designed Make.com Automations to Generate All Content Types
Built multi-step workflows that automatically:

• Generate static post graphics
• Create carousel content and designs
• Produce AI video prompts and video scripts to create AI videos using Creatify and Kling API
• Generate captions + hashtags
• Organize posts for publishing

This removed the need for manual design and creative work.

4. Added a Quality-Control Module
To ensure content met a professional standard, I built:

• A quality-check step using AI to verify tone, accuracy, and design consistency
• Rejection logic (if quality didn't meet the threshold, regenerate the content)

This made the system “self-correcting.”

5. Full Scheduling Automation With Metricool

• Synced Make.com with Metricool for publishing
• Automated content scheduling
• Allowed for multi-platform posting
• Completely removed the need for manual uploads

The Result

• Enabled Guru Labs to save 20+ hours a week on social media content creation across multiple brands
• Consistent, on-brand posts across all brands