Docker Basics: Why Every AI Developer Needs Containers

The Problem Docker Solves

"It works on my machine" is the most expensive phrase in software development. Different Python versions, missing packages, OS differences — all of these cause production failures. Docker packages your entire environment — code, dependencies, Python version, OS settings — into a single container that runs identically everywhere.

For AI apps specifically, this matters because:

• ML dependencies (PyTorch, CUDA) are notoriously version-sensitive
• Cloud platforms (AWS, GCP, Azure) expect containerized apps
• Your teammates need the exact same environment you have

Install Docker

Download Docker Desktop from docker.com. Verify installation:

docker --version
docker run hello-world

Core Docker Concepts

• Image: A read-only template. Like a blueprint. Built from a Dockerfile.
• Container: A running instance of an image. Like a house built from the blueprint.
• Dockerfile: Instructions for building the image. You write this.
• Registry: Where images are stored. Docker Hub is the public one.

Your First Container

# Run Python 3.11 in a container
docker run -it python:3.11-slim python

# Inside the container:
>>> import sys
>>> print(sys.version)  # 3.11.x

# Exit with Ctrl+D
# The container stops when Python exits

docker ps           # List running containers
docker ps -a        # List all containers (including stopped)
docker images       # List local images
docker rm [id]      # Remove a container
docker rmi [image]  # Remove an image
docker logs [id]    # View container logs
docker exec -it [id] bash  # Shell into running container