DevOps & MLOps DVC, AWS & MLflow GitHub Actions
What You’ll Learn
Building a model is not enough — learn how to deploy, monitor, and automate it. This course is the best combination of DevOps and MLOps — Docker, Kubernetes, CI/CD, GitHub Actions, MLflow, DVC, and AWS. A perfect course for both AI Engineers and Software Developers.
Benefits
- Docker — Containerization & Compose
- Kubernetes — Orchestration & Helm
- GitHub Actions — CI/CD Pipelines
- MLflow — Experiment Tracking & Model Registry
- DVC — Data & Model Version Control
- AWS — EC2, S3, SageMaker, Lambda
- Infrastructure as Code (Terraform basics)
- Model Drift Monitoring
Curriculum
🐳 Module 1 — Docker
- Docker Architecture: Engine, Images, Containers, Registry
- Dockerfile: Writing, Building, Optimizing (multi-stage builds)
- Docker Commands: run, exec, ps, logs, inspect, volumes, networks
- Docker Compose: Multi-container apps (Flask + Redis + Postgres)
- Docker Hub: Push, Pull, Private Registries (ECR, GCR)
☸️ Module 2 — Kubernetes
- K8s Architecture: Master Node, Worker Nodes, etcd, API Server
- Objects: Pods, Deployments, Services, ConfigMaps, Secrets, Ingress
- kubectl: Applying manifests, Rollouts, Rollbacks, Port-forwarding
- Helm: Chart creation, Values overriding, Deploying apps
- Horizontal Pod Autoscaling, Resource Limits, Liveness/Readiness probes
- EKS/GKE: Managed Kubernetes basics
⚙️ Module 3 — CI/CD Pipelines
- Git Workflows: Branching strategies, PR reviews, Protected branches
- GitHub Actions: Triggers, Jobs, Steps, Marketplace Actions
- CI Pipeline: Lint → Test → Build → Dockerize → Push
- CD Pipeline: Pull → Deploy to EC2/K8s → Health Check → Notify
- Jenkins: Pipeline as Code (Jenkinsfile), Blue-Green Deployments
- Secrets Management: GitHub Secrets, HashiCorp Vault basics
🤖 Module 4 — MLOps
- MLflow: Tracking Experiments, Logging Metrics/Params/Artifacts
- MLflow Model Registry: Staging → Production promotion workflow
- DVC: Data Versioning, Pipeline stages, Remote Storage (S3)
- FastAPI: Wrapping ML models as REST APIs
- Model Drift Monitoring: Data drift, Concept drift, Evidently AI
- AWS SageMaker: Training Jobs, Endpoints, Model Monitor
- MLflow + GitHub Actions: Automated retraining pipelines
- Project: End-to-end MLOps Pipeline (Train → Track → Deploy → Monitor)
