How Docker Helps You Land a Data Scientist Job Faster

pandas==2.2.3
numpy==2.1.3
scikit-learn==1.5.2
matplotlib==3.9.2
joblib==1.4.2
jupyterlab==4.2.5
fastapi==0.115.4
uvicorn==0.32.0

FROM python:3.12-slim
WORKDIR /work
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt
COPY . .
CMD ["python", "train.py"]

# Using the project image built from your Dockerfile
docker build -t churn-project:1.0 .
docker run --rm -p 8888:8888 -v "$PWD":/work churn-project:1.0 \
  jupyter lab --ip=0.0.0.0 --port=8888 --no-browser --allow-root
# Open the URL with the token printed in the terminal

# compose.yaml — one command for reviewers: docker compose up
services:
  notebook:
    build: .
    command: jupyter lab --ip=0.0.0.0 --port=8888 --no-browser --allow-root
    ports: ["8888:8888"]
    volumes: [".:/work"]

# train.py
import joblib
from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier

X, y = load_iris(return_X_y=True)
model = RandomForestClassifier(n_estimators=100, random_state=42).fit(X, y)
joblib.dump(model, "model.joblib")

# app.py
import joblib
from fastapi import FastAPI
from pydantic import BaseModel

model = joblib.load("model.joblib")
app = FastAPI()

class Flower(BaseModel):
    sepal_length: float
    sepal_width: float
    petal_length: float
    petal_width: float

@app.post("/predict")
def predict(f: Flower):
    x = [[f.sepal_length, f.sepal_width, f.petal_length, f.petal_width]]
    return {"class": int(model.predict(x)[0])}

# Dockerfile CMD for serving:
# CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8000"]
python train.py
docker build -t iris-api:1.0 .
docker run --rm -p 8000:8000 iris-api:1.0 \
  uvicorn app:app --host 0.0.0.0 --port 8000
curl -X POST localhost:8000/predict -H "Content-Type: application/json" \
  -d '{"sepal_length":5.1,"sepal_width":3.5,"petal_length":1.4,"petal_width":0.2}'

.git
.venv
__pycache__/
.ipynb_checkpoints/
data/raw/
*.csv
.env

build:
	docker build -t churn-project:1.0 .

notebook:
	docker compose up

train:
	docker run --rm -v "$$PWD":/work churn-project:1.0 python train.py

api:
	docker run --rm -p 8000:8000 churn-project:1.0 uvicorn app:app --host 0.0.0.0 --port 8000

# Customer churn prediction
## Run in one command
docker compose up   # opens JupyterLab on port 8888

## What is inside
- notebooks/01_eda.ipynb — exploration
- train.py — reproducible training (random_state fixed)
- app.py — FastAPI prediction service
## Results
Recall and precision on held-out data, with limitations