🧠 DP-100 Mini Projects — Azure ML Study & Deployment Toolkit

Welcome to the DP-100 Mini Projects repository! This repo is designed to support hands-on, reproducible learning for the Azure Data Scientist Associate certification, with a focus on project-based mastery, Thai-English datasets, and deployment-ready workflows for classroom and PD use.

---

📅 Weekly Schedule (Updated for April 2025)

Week 1: Environment Setup & Workspace Design

Focus: Azure ML workspace, compute targets, datastores, Git integration
Mini Project: 🔧 Build Your Lab — Deploy a reproducible ML workspace with Git-linked assets and compute targets.

Tasks:

• Create Azure ML workspace and link to GitHub
• Set up compute targets (CPU, GPU, Spark pool)
• Register datastore and upload sample datasets (e.g., Thai rainfall, NYC taxi)
• Configure environments with custom Conda dependencies

---

Week 2: Data Exploration & Experimentation

Focus: AutoML, notebooks, Synapse Spark, MLflow tracking
Mini Project: 📊 Sentiment Explorer — Build and compare AutoML models for Thai/English review data.

Tasks:

• Use AutoML for tabular and NLP tasks (e.g., Thai restaurant reviews)
• Explore data using notebooks and Synapse Spark
• Retrieve features and track experiments with MLflow

---

Week 3: Model Training, Pipelines & Deployment

Focus: Custom training scripts, job config, pipelines, online/batch deployment
Mini Project: 🚀 Smart Deployment — Train a model to predict school attendance trends and deploy endpoints.

Tasks:

• Write and run training scripts with parameterized inputs
• Build pipelines with custom components
• Deploy models to online and batch endpoints with logging

---

Week 4: Language Model Optimization & RAG

Focus: Prompt engineering, prompt flow SDK, RAG, fine-tuning
Mini Project: 🧩 Ask the Archive — Build a RAG-based Q&A bot using Thai educational documents.

Tasks:

• Benchmark a language model from Azure AI Foundry
• Use prompt flow SDK to define chaining logic
• Embed data for RAG using Azure AI Search
• Fine-tune on Thai-English educational dataset

---

🧪 Bonus Week: Evaluation & Extension

Focus: Responsible AI, model monitoring, classroom integration
Mini Project: 🛡️ Model in the Wild — Monitor a deployed model and simulate real-world drift.

Tasks:

• Evaluate fairness and explainability with Responsible AI dashboard
• Set up model monitoring for endpoints
• Draft classroom activities using simplified project versions

---

🗂️ Repo Structure

DP-100-Mini-Projects/
├── Week1_BuildYourLab/
├── Week2_SentimentExplorer/
├── Week3_SmartDeployment/
├── Week4_AskTheArchive/
├── BonusWeek_ModelInTheWild/
└── README.md

---

## 🏷️ Repo Tags

This project touches on the following key areas:

- `azure-machine-learning`
- `dp-100`
- `data-science`
- `automl`
- `mlflow`
- `rag`
- `promptflow`
- `azure-ai`
- `responsible-ai`
- `thai-language`
- `nlp`
- `education`
- `project-based-learning`
- `reproducible-research`
- `classroom-deployment`
- `ai-in-education`