The Challenge
MCRDSE's content operation faced common scaling bottlenecks: Content creation was manual and time-intensive across 5+ platforms No systematic way to repurpose long-form content (YouTube videos) into short-form (Instagram, LinkedIn) Brand voice inconsistency across different content creators and platforms No centralized content pipeline - everything lived in scattered documents and spreadsheets Email newsletters required manual compilation of the week's content SEO blog posts required research, writing, editing, and formatting - each taking hours
What We Did
Core Infrastructure
The 5-Agent AI Pipeline
Each content piece flows through specialized AI agents, each optimized for a specific task:
Agent 1 - Research & Analysis Ingests raw materials: YouTube transcripts, URLs, uploaded documents, topic briefs Performs competitive analysis on the topic Extracts key themes, data points, and quotable insights Produces a structured research brief for downstream agents Model: GPT-4o (high reasoning capability needed)
Agent 2 - Content Strategy Takes the research brief and determines optimal content angles Maps content to platform-specific formats (long-form blog, carousel, reel script, LinkedIn post, email section) Assigns brand voice parameters per platform Creates content outlines with section-by-section briefs Model: Claude (strong at strategic/structural thinking)
| Layer | Technology |
|---|---|
| Dashboard | Next.js (App Router) + TypeScript |
| Workflow Engine | n8n (self-hosted on Hetzner/Coolify) |
| Content Database | Notion (multi-database setup) |
| AI Models | OpenAI GPT-4o, Anthropic Claude, Google Gemini |
| Video Processing | YouTube Data API + transcript extraction |
| Image Generation | DALL-E 3 / Midjourney API |
| Brevo (Sendinblue) API | |
| Analytics | Custom tracking via Notion + dashboard |
| Deployment | Docker on Hetzner via Coolify |
The Results
End-to-End Content Automation Capability
This is not a single chatbot or a simple automation. It is a complete content production system - a factory that takes raw inputs and produces platform-ready content across 5+ channels with brand consistency, quality gates, and performance tracking.
Multi-AI Orchestration
The 5-agent pipeline demonstrates sophisticated AI orchestration - using different models for different cognitive tasks (research, strategy, generation, review, distribution) rather than forcing one model to do everything. This approach produces higher quality output at lower cost.
Scalable Architecture
Key Takeaway
End-to-End Content Automation Capability
This is not a single chatbot or a simple automation. It is a complete content production system - a factory that takes raw inputs and produces platform-ready content across 5+ channels with brand consistency, quality gates, and performance tracking.
Multi-AI Orchestration
The 5-agent pipeline demonstrates sophisticated AI orchestration - using different models for different cognitive tasks (research, strategy, generation, review, distribution) rather than forcing one model to do everything. This approach produces higher quality output at lower cost.
Scalable Architecture