Logo
Back to blog

The Codebase Approach to MBA Studies

•8 min read
mba

Throughout my MBA, I've tried every organizational method under the sun. Apple Notes, Notion, various native file structures—you name it, I've probably given it a shot. After 18 months and eight quarters, I settled on something almost embarrassingly simple: a plain structured directory.

Here's what it looks like:

School/
├── 01 Fall 2023/
│   ├── Class 1/
│   │   ├── Week 1/
│   │   ├── Week 2/
│   │   └── ...
│   └── Class 2/
│       ├── Week 1/
│       ├── Week 2/
│       └── ...
├── 02 Winter 2024/
│   └── [Similar structure]
└── 99 Admin/
    ├── Scholarships/
    ├── Financing/
    └── Career Dev/

Why did this win? It's scalable, simple, and syncs effortlessly via iCloud across all my devices. No learning curve, no proprietary formats, just folders and files.

The AI Context Problem

Like most modern students, AI has become indispensable to my learning. Personal tutor, writing assistant, concept explainer—the list goes on. But there's been one consistently annoying friction point: context setup.

Every single time I wanted to work on something, I'd go through the same ritual. Open the chat, locate my files, copy relevant sections, explain what class this is for, remind the AI what we covered last week. Not a huge deal on its own, but multiply that by hundreds of sessions? The time adds up.

It felt like reintroducing myself to a colleague with amnesia every time we needed to work together.

My Parallel Life in Code

On the side—in what limited free time an MBA student has—I've been teaching myself to code. Built a few small apps, dabbled in dozens of libraries and frameworks, tried to understand SDLC and engineering best practices. And naturally, AI has been just as crucial here.

I went through the whole tooling journey that any developer in 2025 knows well: V0, Replit, Bolt.new, Copilot, Windsurf, Cursor, Claude Code, Gemini CLI. After extensive experimentation, I landed on Claude Code as my primary agentic coding tool.

What sold me? It has access to my entire codebase, can search the web when needed, leverages MCPs for advanced functionality, and writes code directly in my project. But the killer feature was how it understood my project holistically. Unlike Windsurf or Cursor, Claude Code felt faster, performed better, and just got what I was building.

The Epiphany

Eight quarters into my MBA, it hit me: a codebase is just a structured folder hierarchy. It contains files of different types, can be tracked through version control, has a clear organization system. Sound familiar?

My meticulously organized MBA folders were already a codebase. I just wasn't treating them like one.

Making the Connection

The implementation was almost anticlimactic in its simplicity. I opened my School directory in VSCode and fired up Claude Code. Suddenly, my AI assistant could see everything—lecture notes, Excel models, case studies, all in their proper context.

Want to work on that private equity case? Claude can reference the LBO model from Week 3, the valuation framework from last Tuesday, and the industry analysis from my Strategy class. No manual context-setting required.

But here's where it gets really interesting: VSCode handles everything. PDFs, Word docs, PowerPoints, Excel files—all accessible in one interface. Instead of juggling Adobe Acrobat, Microsoft Office, and a browser, I have a single workspace where I can do all my work AND collaborate with my AI tutor.

Why This Matters

This isn't just about saving time on context-switching (though that alone makes it worthwhile). It's about creating a learning environment that matches how AI actually works best—with complete visibility into your knowledge base.

Consider what this enables:

  • Claude can identify patterns across courses I hadn't noticed
  • Version control means I can track how my understanding evolves
  • Cross-referencing happens automatically ("This NPV calculation relates to what you learned about in Corporate Finance Week 4")
  • My entire MBA becomes a searchable, interconnected knowledge graph

The Bigger Picture

What started as a simple productivity hack revealed something more profound. The distinction between "technical" and "non-technical" work is becoming meaningless. The same principles that make codebases manageable—structured organization, version control, systematic documentation—apply equally to any knowledge work.

We're moving toward a world where the ability to structure information for both human and machine consumption isn't just useful—it's essential. By treating my MBA materials like a codebase, I didn't just solve an organizational problem. I created a learning system that scales with AI capabilities.

Looking Forward

As I enter my final quarters, this system continues to evolve. I'm exploring Git for document version control, automated analysis of my academic performance, and custom scripts to surface insights from my coursework.

But the real lesson is this: sometimes the most powerful innovations come from applying existing tools in unexpected ways. Software developers solved the problem of managing complex, interconnected information systems years ago. By borrowing their approach, I've transformed not just my workflow, but my entire relationship with learning.

The future of education isn't about building new tools—it's about recognizing that the tools we need already exist. We just need to think differently about what a "codebase" can be.

Have you found unexpected applications for developer tools in your work? What would change if you treated your professional materials like a software project? I'd love to hear how others are bridging these worlds.