Machine Learning - Production Overview

Deployment patterns

• Batch scoring to a store (offline).
• Online low-latency APIs (autoscaling, caching, fallbacks).
• On-device/Edge (CoreML/TFLite/ONNX).

Simple FastAPI serving

# pip install fastapi uvicorn scikit-learn
from fastapi import FastAPI
from pydantic import BaseModel
import joblib, numpy as np

class Item(BaseModel):
    features: list

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

@app.post("/predict")
def predict(item: Item):
    X = np.array([item.features])
    y = model.predict(X).tolist()
    return {"prediction": y}