مدرب MLOps لتقديم دورة مكثفة

أريد MLOps developer trainer لدي خبرة لا تقل عن خمس سنوات فى هذا المجال لتقديم دورة مكثفة مدتها شهرين لمساعدتى على تعلم هذا المجال ، يمكن التعديل على ترتيب المحتوى أو إضافة موضوعات حسب رؤية المدرب ولا يمكن حذف اي موضوع.

محتوى الدورة المبدئى:

**Week 1: MLOps Overview + Production Thinking**

- Train → package → deploy → monitor (end-to-end view)

- Define a capstone: an ML/DL model served via API + monitoring + CI/CD

**Week 2: Python Packaging + Project Structure**

- Production-grade project layout

- Configuration, logging basics

- Testing mindset

**Week 3: REST API Development (FastAPI) + Docker**

- Build a prediction API (request/response schemas)

- Containerize with Docker

- Basic reliability: input validation + error handling

**Week 4: Test-Driven Development (TDD) with pytest**

- Unit tests for preprocessing/inference

- Integration tests for API

- Regression testing strategy

**Week 5: Training Pipeline & Experiment Setup**

- Dataset loading + preprocessing

- Baselines: Linear Regression / XGBoost / simple DL model (MLP)

- Use MLflow/W&B to track experiments and artifacts

-SEMI PROJECT

---

**Week 6: Evaluation & Model Selection**

- Proper train/val/test splits

- Metrics for classification/regression

- Compare experiments and select the best run

**Week 7: Experiment Tracking + Model Registry Concepts**

- Promote runs from dev → staging → production (conceptually)

- Artifact management (models, scalers, preprocessing)

- Multi-model comparison workflow

Week 8 : End-to-End CI/CD + Data Versioning + IaC**

- CI: GitHub Actions for tests/build/lint

- DVC for data versioning

- Terraform for Infrastructure as Code

- Continuous training triggers (concept + setup)

- Introduction to LLM & RAG

- final project using all what we learned using LLM