Tell how you integrated machine learning models into the server part of an application at your previous workplace.

At my previous job, I integrated machine learning models into the server-side through REST API. The model was trained separately, then exported (for example, in pickle or ONNX format). An endpoint was created on the server that accepted input data, passed it to the model, and returned predictions.

Example in Python using Flask:

from flask import Flask, request, jsonify
import pickle

app = Flask(__name__)

# Load the model
with open('model.pkl', 'rb') as f:
    model = pickle.load(f)

@app.route('/predict', methods=['POST'])
def predict():
    data = request.json['data']
    prediction = model.predict([data])
    return jsonify({'prediction': prediction[0]})

if __name__ == '__main__':
    app.run()

This approach allows separating training from the service and easily scaling the application.