🧠 AI/ML Engineering with Python - Examples

This monorepo contains a collection of modular, end-to-end projects that demonstrate best practices for AI/ML engineering — from experimentation to production — with a focus on:

• MLOps & LLMOps patterns
• Real-world infrastructure integration
• CI/CD pipelines
• Evaluation, monitoring, and observability

These projects are designed to support a forthcoming book on AI/ML Engineering, and mirror the full lifecycle of modern ML and LLM systems.

---

📚 What This Repo Covers

Each project explores different slices of the AI/ML engineering stack, including:

• ✅ Training & Experimentation: with Ray, Kedro, classic ML, notebooks
• 🧠 LLM Apps & Agents: Bedrock, CrewAI, LangGraph, RAG workflows
• 🚀 Deployment & Inference: via Ray Serve, FastAPI, Bedrock, Lambda
• 📈 Monitoring & Observability: CloudWatch, Evidently AI, MLflow
• 🧪 Evaluation: model metrics, agent output quality, prompt tests
• 🔁 CI/CD Workflows: GitHub Actions, reusable runners, CT/CI patterns
• 🧰 Infrastructure: Terraform/CDK, Docker, Python envs

---

🏗️ Monorepo Structure

projects/
├── ray-ml-pipeline/           # Ray Datasets, Train, Serve
├── kedro-ml-pipeline/         # Modular pipeline + optional MLflow
├── bedrock-rag-agent/         # Full-stack LLM app (client + server)
├── agent-frameworks/          # CrewAI, LangGraph, Pydantic-AI
├── classic-ml/                # Traditional models + CI
├── observability-pipelines/  # Lambda, CloudWatch, metrics
shared/                        # Common utilities across projects
evaluation/                    # Shared evaluation logic and runners
infra/                         # Optional shared Terraform/CDK modules
.github/                       # CI/CD workflows