Expert-Driven Development (EDD) is a methodology that teaches domain experts to build institutional software using AI as scaffolding, not as an end-state dependency. It is designed for military instructors, staff NCOs, officers, and DoW civilians who have real problems to solve but no coding background. This repository contains the complete training curriculum (five courses covering every role from first-time AI users to experienced builders to supervisors), the Standard Operating Procedure, an interactive toolkit, and downloadable templates. Everything here is open-source, unclassified, and ready to use.
Domain experts understand their problems better than anyone, but they cannot build the solutions themselves. Traditional software development requires contractors, months of lead time, and significant budgets. The people closest to the problem remain disconnected from the tools that could solve it.
EDD uses AI as scaffolding to let domain experts build, document, and maintain their own tools on approved platforms. The tools outlast the AI -- they are standard applications running on standard infrastructure, not black-box AI products. One instructor built a system serving 440 users in 30 hours at zero cost.
| Layer | Name | Purpose |
|---|---|---|
| 1 | Train | Develop AI-fluent builders through structured coursework |
| 2 | Build | Create institutional tools using AI-assisted development |
| 3 | Document | Ensure continuity, replication, and knowledge transfer |
| 4 | Scale | Spread capability organically across the organization |
| Course | Duration | Audience | Outcome |
|---|---|---|---|
| AI Fluency Fundamentals | 2 hours | All personnel | Understand six 201 skills, recognize the jagged frontier, map AI into workflow |
| Builder Orientation | 2 hours | Aspiring builders | Build a working prototype, apply task decomposition and iterative refinement |
| Platform Training | 4 hours | Builders | Build 3 complete tools on Power Platform using centaur and cyborg patterns |
| Advanced Workshop | 4 hours | Experienced builders | Map the frontier, build verification protocols, teach others |
| Supervisor Orientation | 30 minutes | Leadership | Create permission culture, evaluate proposals, understand apprentice problem |
This is a GitHub Pages site served from the /docs directory. All training materials, resources, and tools are available directly through the live site.
-
Visit the Training Site -- Access all courses, resources, and tools at jeranaias.github.io/ExpertDrivenDevelopment
-
Start with Course 1 -- Complete AI Fluency Fundamentals to understand the six core skills and recognize the jagged frontier.
-
Follow the Builder Pathway -- Progress through Courses 2-4 to develop institutional tools, or direct leadership to Course 5.
-
Use the Resources -- Access the prompt library, FAQ, templates, and interactive toolkit from the site.
- MCD Tutoring Application -- 440 users, 30 hours to build, $0 cost, ATO approved
- DonDocs (Marine Coders) -- 300 unique users/day on MCEN, MIU recognized
- Tanaghum -- ILR-calibrated Arabic listening lessons, built in 2 days, in use by MLIs
- Harakat -- Arabic diacritization system, 0.68% DER
- Mutawazin -- Arabic typing proficiency trainer
Universal Pathway (All Personnel):
- Course 1: AI Fluency Fundamentals → Course 5: Supervisor Orientation (leadership)
Builder Pathway:
- Course 1: AI Fluency Fundamentals → Course 2: Builder Orientation → Course 3: Platform Training → Course 4: Advanced Workshop
The repository includes student-facing versions of Courses 1-4, designed to work as both self-paced courses and in-person companion materials. These simplified versions focus on hands-on exercises and are located at docs/courses/student/. There is no student version of Course 5 (Supervisor Orientation), as it is leadership-focused and delivered as a briefing.
ExpertDrivenDevelopment/
├── .github/ # Community files
│ ├── ISSUE_TEMPLATE/ # Issue templates
│ ├── CONTRIBUTING.md # Contribution guidelines
│ └── PULL_REQUEST_TEMPLATE.md
├── docs/ # GitHub Pages site (live at jeranaias.github.io/ExpertDrivenDevelopment)
│ ├── index.html # Site home page
│ ├── about.html # About EDD methodology
│ ├── courses/ # Training course pages
│ │ ├── index.html # Course catalog
│ │ ├── ai-fluency.html # Course 1: AI Fluency Fundamentals
│ │ ├── orientation.html # Course 2: Builder Orientation
│ │ ├── platform.html # Course 3: Platform Training
│ │ ├── advanced.html # Course 4: Advanced Workshop
│ │ ├── supervisor.html # Course 5: Supervisor Orientation
│ │ └── student/ # Student-facing course materials
│ │ ├── index.html # Student course catalog
│ │ ├── ai-fluency.html # Student version of Course 1
│ │ ├── orientation.html # Student version of Course 2
│ │ ├── platform.html # Student version of Course 3
│ │ └── advanced.html # Student version of Course 4
│ ├── resources/ # Supporting resources
│ │ ├── index.html # Resources home
│ │ ├── faq.html # Frequently asked questions
│ │ ├── prompt-library.html # Curated prompt templates
│ │ └── templates.html # Project templates
│ ├── sop/
│ │ └── index.html # Standard Operating Procedure
│ ├── toolkit/
│ │ └── index.html # Interactive EDD Toolkit
│ ├── css/
│ │ └── style.css # Site styles
│ ├── pdf/ # Downloadable PDFs
│ └── templates/ # Markdown templates
├── LICENSE # MIT License
├── README.md
└── SECURITY.md # Security and classification policy
Contributions are welcome from all DoW personnel. Please read the Contributing Guide before submitting issues or pull requests.
All contributions must be UNCLASSIFIED. Do not submit CUI, PII, or classified material under any circumstances.
All materials in this repository are UNCLASSIFIED // Distribution Unlimited. No CUI, PII, or classified information is permitted in any file, issue, pull request, or discussion. See SECURITY.md for details.
This project is licensed under the MIT License.
SSgt Jesse C. Morgan Marine Corps Detachment, Presidio of Monterey Defense Language Institute Foreign Language Center jesse.morgan@dliflc.edu