deeploi 0.2.0

One-line deployment for trained tabular ML models

Deeploi — Deploy ML Models in Seconds 🚀

Turn your trained tabular ML model into a production-ready API with a single line of code. No DevOps. No boilerplate. No headaches.

Why Deeploi?

• Instant API: Serve your scikit-learn or XGBoost model in one command.
• Zero Config: No YAML, no Docker, no cloud lock-in.
• Local-First: Run on your laptop, server, or VM.
• Built for ML work: Focus on the model, not the serving stack.

Install

pip install deeploi

Fastest Start

from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
from deeploi import deploy

iris = load_iris(as_frame=True)
X, y = iris.data, iris.target
model = RandomForestClassifier(random_state=42).fit(X, y)

deploy(model)

Your model is now live at http://127.0.0.1:8000.

Most Important Ways to Access Deeploi

1. Open the dashboard

Open this in your browser:

http://127.0.0.1:8000/

The dashboard gives you a quick way to inspect metadata, test endpoints, and send prediction requests.

2. Call the prediction API directly

curl -X POST http://127.0.0.1:8000/predict \
  -H "Content-Type: application/json" \
  -d '{
    "records": [
      {
        "sepal length (cm)": 5.1,
        "sepal width (cm)": 3.5,
        "petal length (cm)": 1.4,
        "petal width (cm)": 0.2
      }
    ]
  }'

Core endpoints:

• POST /predict
• POST /predict_proba
• GET /meta
• GET /health

3. Work with a reusable package object

from deeploi import package

pkg = package(model)
preds = pkg.predict(X.head())
pkg.serve(port=8000)

Use this when you want a Python object you can predict with, save, and serve explicitly.

4. Save and reload a model artifact

from deeploi import package, load

pkg = package(model)
pkg.save("artifacts/iris_rf")

loaded = load("artifacts/iris_rf")
loaded.serve(port=8000)

Use this when you want a portable artifact directory you can reload later.

Docker Artifact Generation

Use Docker generation when you want a saved artifact that is easy to move onto another machine or run in a container.

from deeploi import package

pkg = package(model)
pkg.save("artifacts/iris_rf", generate_docker=True)

Then build and run it:

cd artifacts/iris_rf
docker build -t iris-model .
docker run --rm -p 8000:8000 iris-model

Who is Deeploi for?

• Data scientists who want to share a model quickly
• ML engineers who need fast local serving for tabular models
• Teams that want a lightweight path from notebook to API

Learn More

See DOCS.md for:

• advanced deploy, package, and load usage
• environment-based authentication
• Docker artifact generation
• artifact layout
• error handling and testing examples

License

MIT — see LICENSE.