Case Study: The AI Cortex

The Problem: The AI Knowledge Gap

The field of AI Engineering is exploding, but high-quality, structured educational content often remains fragmented or locked behind expensive paywalls. Aspiring engineers face a steep learning curve, struggling to connect foundational theories like the Transformer architecture with the practical realities of deployment, ethics, and MLOps. The challenge was clear: could I build a comprehensive, interactive, and fully open-source curriculum that would serve as a definitive guide for the next generation of AI engineers?

The Solution: An Interactive, Living Textbook

The answer is The AI Cortex, a complete educational platform designed from the ground up to be a living, breathing textbook. It goes beyond static text by integrating hands-on projects, interactive diagrams, and a deeply interconnected knowledge graph structure.

The core innovation is a pedagogically-sound architecture that guides learners from foundational concepts to production-ready systems. The platform includes "PromptCraft," a fully functional AI application that runs 100% locally, allowing students to experiment with prompt engineering without API keys or costs. This bridges the critical gap between theoretical knowledge and practical, real-world skill.

The Architecture & Tech Stack

I architected a modern, maintainable, and highly performant system using a JAMstack approach, with a focus on automation and developer experience.

The Content Engine (Python & MkDocs)

The entire course, with over 50 interconnected technical notes, is written in Markdown and built using MkDocs, a powerful Python-based static site generator. The Material for MkDocs theme was heavily customized with CSS to create a professional, branded learning environment.

The Interactivity Layer (Mermaid.js)

Dozens of complex concepts are brought to life with custom, interactive diagrams built with Mermaid.js. A custom JavaScript initializer (mermaid-init.js) ensures diagrams are theme-aware and dynamically render correctly in both light and dark modes.

The CI/CD Pipeline (GitHub Actions & Vercel)

The project is hosted in a private GitHub repository. A GitHub Actions workflow defines a robust Continuous Integration/Continuous Deployment (CI/CD) pipeline. On every push to the main branch, the action automatically:

1. Sets up a Python environment.
2. Installs all dependencies from a Pipfile.
3. Runs the mkdocs build command.
4. Deploys the resulting static site directory to Vercel.

The Deployment Platform (Vercel)

The final static site is hosted on Vercel's global Edge Network, providing exceptional performance, automatic SSL, and a professional custom subdomain (cortex.behailu.tech). The entire deployment process is fully automated.

The Outcome: A Complete Educational Ecosystem

The AI Cortex successfully demonstrates that it's possible to create a world-class educational resource that is both deeply technical and entirely open-source. The project itself became a lesson in modern DevOps and platform engineering, solving real-world deployment challenges to deliver a fast, reliable, and globally accessible learning platform.

---

Tech Stack

Python MkDocs Material for MkDocs Mermaid.js JavaScript CSS YAML CI/CD GitHub Actions Vercel