[New] Ultimate Docker Bootcamp for ML, GenAI and Agentic AI

Name: [New] Ultimate Docker Bootcamp for ML, GenAI and Agentic AI
Rating: 4.7 (307 reviews)

Master Docker for real-world AI & ML workflows — Dockerfiles, Compose, Docker Model Runner, Model Context Protocol (MCP)

Role Play

Created byGourav J. Shah, School of Devops

Last updated 7/2025

English

What you'll learn

Run and manage Docker containers tailored for AI/ML workflows
Containerize Jupyter notebooks, Streamlit dashboards, and ML development environments
Package and deploy Machine Learning models with Dockerfile
Publish your ML Projects to Hugging Face Spaces
Push and pull images from DockerHub and manage Docker image lifecycle
Apply Docker best practices for reproducible ML research and collaborative projects
LLM Inference with Docker Model Runner
Setup Agentic AI Workflows with Docker Model Context Protocol (MCP) Toolkit
Build and Deploy Containerised ML Apps with Docker Compose

Course content

6 sections • 45 lectures • 6h 4m total length

Why and How Docker is important for Machine Learning / Artificial Intelligence8:47
Why and How Docker is important for Machine Learning / Artificial Intelligence6:33
Docker in the world of LLMs and Agentic AI5:05
Download the Slides Deck0:01
Installing and validating Docker Desktop8:15
Setting up tools and environment for this Course5:22
Explaining Docker to a Skeptical Stakeholder
Quick Quiz: Why Docker Matters for AI/ML

Project - Setup ML Dev Environment with Docker - MLFlow and Jupyter2:38
Set up an ML development environment with Docker, launch Jupyter Lab notebooks, and use MLflow to track experiments, configurations, hyperparameters, features, and evaluation metrics.
Docker Concepts - Images, Container, Registry, Repository. Pulling Images11:38
Launch MLflow in a Docker container by pulling registry images, understanding images, containers, registries, and repositories, and exploring image layering for portable, consistent environments.
Launch, Analyse and Connect to MLFlow Container9:06
Pull the mlflow image from the GitHub container registry, launch a container with port mapping to expose mlflow's web UI on the host, and monitor the container's lifecycle and events.
Container Operations - Common options, Detaching, Listing, Managing Containers11:16
Launch JupyterLabs Notebook Envioronment with a Volume shared with Host9:28
Writing and executing a simple ML Project with Container Hosted Jupyter Notebook12:00
Connect Notebook with MLFlow Container for Experiment Tracking10:19
Project Guide -Build it Yourself5:23
Lab Review: Running MLflow & Jupyter with Docker

Project Nebula - Containerize Tech Stack Advisor ML App and Host it on Hugging F4:05
Mission Statement3:43
Build and test the ML Project and Train the Model10:28
Why and how to build Container Images Manually First ?6:06
Learn to containerize a machine learning app by performing a manual test build, then switch to a Dockerfile workflow with a base image, copy code, install dependencies, and test run.
Building a container image step by step using the Imperative Approach13:23
Buidling and Testing the Image using Dockerfile8:08
Analysing, Tagging and Publishing Container Images9:15
How to write Dockerfile ? Instructions Quick Dive13:10
Deploy and Host Containerized App to Hugging Face Spaces12:24
Project Guide -Build it Yourself3:43
Quick Quiz: Containerizing ML Apps with Dockerfile
Building a Local ML Stack with Docker Compose

Project - Build and Deploy House Price Predictions ML App in Dev with Docker Co4:51
Mission Statement2:00
Understanding the Application Stack and the ML Workflow10:19
Build a Docker Compose stack to onboard dev teams, simulating production for a house price predictor with a Streamlit front end and a FastAPI back end, all logged in MLflow.
Automate MLFlow Launch with Code by writing Compose Spec , Learn Compose Syntax12:51
Automate mlflow deployment with docker compose by converting docker run to a YAML compose spec, defining mlflow as a service, ports, image, and command for a shared dev workflow.
Run the Data Processing, Feature Engineering and Model Training Pipeline for Hou10:54
Learn to build and run an ml pipeline—data processing, feature engineering, and model training—for house price prediction, logged with mlflow and packaged for dockerized fast api and streamlit deployment.
Composing FastAPI and Streamlit Apps with Multi Service Compose Spec16:55
Connecting Services using DNS Based Service Discovery offered by Docker Compose12:49
Project Guide - Build it Yourself2:34

Project - Integrate LocalGPT App with Locally Running LLM using Docker Model Run2:41
Mission Statement1:50
What is Docker Model Runner ? How to Set it up with Docker Desktop ?6:06
Exploring Docker Model Runner - Pull a LLM Model from Gen AI Catalogue and Run i13:34
pull AI models from Docker Hub and Hugging Face using the Docker model runner, launch a small M2 or Gamma three model locally, and test open AI type endpoints.
Lanching LocalGPT App with Docker Model Runner with OpenAI Compatible Connection9:39
Configuring Docker Model Runner as a Provider to Compose14:48
Learn to configure a Docker model runner as a provider in Docker Compose to connect a local language model, enabling dynamic model switching via environment variables and without code changes.
Quick Quiz: Running Local LLMs with Docker Model Runner

Project - Explore Docker MCP Toolkit2:28
Mission Statement1:26
What is Model Context Protocol (MCP) and how it sets the foundation for Agentic11:58
Getting started with Docker MCP Toolkit and MCP Catalogue with Gordon AI10:31
Using Filesystem MCP Server6:41
Configure the docker mcp toolkit to expose the file system as an mcp server, connect with gordon, manage containers via docker cli, and generate a python script to locate duplicates.
Securely connecting to GitHub MCP Server and using Read Only Access10:50
Connect to a GitHub MCP server through Docker Gordon, configure a read-only access token, and explore repositories, commits, and basic issue creation, highlighting secure token management.
Auto generating Code, Revision Controlling it and Pushing it toGitHub with MCP S8:34
Quick Quiz: Connecting LLMs to Tools with Docker MCP

Requirements

Basic understanding of Python — you don’t need to be an expert, but you should be comfortable running scripts or working in notebooks.
Familiarity with Machine Learning concepts — knowing what a model is, and having used libraries like scikit-learn, pandas, or TensorFlow will help.
Laptop with Docker/Rancher installed — we’ll walk you through setting up Docker Desktop for Windows, macOS, or Linux.
A GitHub account (recommended) — for accessing project code and pushing your own.
Curiosity to build real-world AI/ML projects with Docker — no prior Docker experience is required!

Description

Welcome to the ultimate project-based course on Docker for AI/ML Engineers.

Whether you're a machine learning enthusiast, an MLOps practitioner, or a DevOps pro supporting AI teams — this course will teach you how to harness the full power of Docker for AI/ML development, deployment, and consistency.

What’s Inside?

This course is built around hands-on labs and real projects. You'll learn by doing — containerizing notebooks, serving models with FastAPI, building ML dashboards, deploying multi-service stacks, and even running large language models (LLMs) using Dockerized environments.

Each module is a standalone project you can reuse in your job or portfolio.

What Makes This Course Different?

Project-based learning: Each module has a real-world use case — no fluff.
AI/ML Focused: Tailored for the needs of ML practitioners, not generic Docker tutorials.
MCP & LLM Ready: Learn how to run LLMs locally with Docker Model Runner and use Docker MCP Toolkit to get started with Model Context Protocol
FastAPI, Streamlit, Compose, DevContainers — all in one course.

Projects You'll Build

Reproducible Jupyter + Scikit-learn dev environment
FastAPI-wrapped ML model in a Docker container
Streamlit dashboard for real-time ML inference
LLM runner using Docker Model Runner
Full-stack Compose setup (frontend + model + API)
CI/CD pipeline to build and push Docker images

By the end of the course, you’ll be able to:

Standardize your ML environments across teams
Deploy models with confidence — from laptop to cloud
Reproduce experiments in one line with Docker
Save time debugging “it worked on my machine” issues
Build a portable and scalable ML development workflow

Who this course is for:

Data Scientists and ML Engineers who want to productionize their workflows
AI/ML Practitioners looking to containerize and deploy models easily
DevOps Engineers supporting AI teams and looking to build ML-ready pipelines
AI Hobbyists and Learners who want to run LLMs or dashboards locally using containers
Anyone tired of “it works on my machine” issues in ML environments

[New] Ultimate Docker Bootcamp for ML, GenAI and Agentic AI

What you'll learn

Explore related topics

Course content

Introduction8 lectures • 34min

Launch and Operate ML Dev Environments with Docker9 lectures • 1hr 12min

Packaging ML Apps as Container Images with Dockerfiles12 lectures • 1hr 24min

Simualting Production Grade ML Systems in Dev with Docker Compose8 lectures • 1hr 13min

Running LLMs Locally with Docker Model Runner7 lectures • 49min

Exploring Model Context Protocol with Docker MCP Toolkit8 lectures • 52min

Requirements

Description

Who this course is for: