Secure AI Platform – Meta Repository¶

Welcome to the Secure AI Platform (SAI Platform) documentation hub. This site is powered by MkDocs and serves as the single source of truth for all documentation, architecture, and governance resources that guide the Secure AI Platform project.

🌐 Overview¶

The Secure AI Platform is an open-source initiative that demonstrates how to build, deploy, and secure AI systems in a real-world, production-style environment. It combines modern MLOps practices with DevSecOps principles to showcase the secure design of AI infrastructure, APIs, and agents.

This repository — sai-platform-meta — acts as the governance and documentation layer of the entire platform. It defines how each module is built, tested, secured, and documented.

🧭 What You’ll Find Here¶

Category	Description
Architecture	High-level platform designs, diagrams, and specifications.
Narrative	Plain-language explanations, project stories, and educational write-ups.
Technical	In-depth engineering references, configuration notes, and security baselines.
Workflows & Templates	Shared GitHub Actions, issue templates, and contribution standards.

🧩 Platform Modules¶

Each major component of the Secure AI Platform lives in its own repository and integrates through shared standards defined here.

Module	Purpose
sai-platform-meta	Central documentation, governance, CI/CD templates.
sai-platform-infra	Local and cluster infrastructure (Vault, Loki, Grafana, OTel, etc.).
sai-ml-foundations	Machine learning pipelines and experiment tracking.
sai-inference-api	Secure model serving API with JWT and RBAC.
sai-agent-secops	Secure AI agents and automation for SOC workflows.
sai-adversarial-lab	Adversarial testing, red teaming, and model robustness validation.
sai-mlops-pipeline	Central CI/CD, scanning, and signing pipelines.

For an architectural overview of how these modules connect, see Platform Overview.

⚙️ Getting Started Locally¶

You can set up the local environment entirely with Python — no system-level dependencies required.

# Clone the repository
git clone https://github.com/EODWeber/sai-platform-meta.git
cd sai-platform-meta

# Set up local virtual environment and dependencies
make setup

# Run all pre-commit checks
. .venv/bin/activate
pre-commit run -a

# Serve documentation locally
make docs

When you’re ready, open http://127.0.0.1:8000 in your browser to explore the site.

🛠 Documentation Structure¶

docs/
├── index.md                 # You are here
├── architecture/            # Technical structure and diagrams
│   └── PLATFORM_OVERVIEW.md
├── narrative/               # Project stories and accessible explanations
│   ├── PLATFORM_STORY.md
│   └── REPO_STORY.md
└── tech/                    # Deep dives, baselines, and configuration notes

Each section can be read independently, or as part of the full story:

Start with Platform Story for an accessible overview.
Explore Architecture Overview for technical depth.
Drill down into module-specific docs as they’re added.

🧱 Building and Publishing Docs¶

This site is built with MkDocs Material.

# Serve locally
make docs

# Build static site
.venv/bin/mkdocs build

Docs are published automatically through GitHub Actions to:

https://<your-username>.github.io/sai-platform-meta/

If you’re contributing or running locally, the only requirement is Python 3.11+.

🧩 Contributing¶

This project follows a security-first and open documentation model. Before submitting changes:

Run pre-commit run -a to lint and scan your code.
Ensure documentation follows the existing format and voice.
Include or update relevant architecture diagrams when adding new modules or workflows.

🧠 Learn More¶

Platform Story – High-level explanation in plain language.
Architecture Overview – System structure and relationships.
Meta Repo Story – Why this repository exists and how to use it.