Azure Machine Learning Docs

Home

How to Consume Azure Machine Learning Models from a Web Application

This guide shows you how to call an Azure Machine Learning (AML) model endpoint from a web application using both Python (Flask) and C# (ASP.NET Core). You’ll learn how to secure the request with an authentication token, send data, and handle the response.

Prerequisites

Flask (Python) Example

Tip: Install the required packages with pip install flask requests. 
# app.py
import os
import json
import requests
from flask import Flask, request, jsonify

app = Flask(__name__)

AML_ENDPOINT = os.getenv('AML_ENDPOINT')
AML_KEY = os.getenv('AML_KEY')
HEADERS = {
    'Content-Type': 'application/json',
    'Authorization': f'Bearer {AML_KEY}'
}

@app.route('/predict', methods=['POST'])
def predict():
    # Forward the incoming JSON payload to AML endpoint
    payload = request.get_json()
    resp = requests.post(AML_ENDPOINT, headers=HEADERS, json=payload)
    if resp.status_code != 200:
        return jsonify({'error': resp.text}), resp.status_code
    return jsonify(resp.json())

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

ASP.NET Core (C#) Example

Tip: Use dotnet add package Microsoft.Extensions.Http to ensure HttpClientFactory is available. 
// Startup.cs (or Program.cs for .NET 6+)
using System.Net.Http.Headers;

var builder = WebApplication.CreateBuilder(args);
builder.Services.AddHttpClient("AMLClient", client =>
{
    client.BaseAddress = new Uri(builder.Configuration["AML:Endpoint"]);
    client.DefaultRequestHeaders.Accept.Add(new MediaTypeWithQualityHeaderValue("application/json"));
    client.DefaultRequestHeaders.Authorization = new AuthenticationHeaderValue("Bearer", builder.Configuration["AML:Key"]);
});

var app = builder.Build();

app.MapPost("/predict", async (HttpContext http, IHttpClientFactory httpClientFactory) =>
{
    var requestBody = await new StreamReader(http.Request.Body).ReadToEndAsync();
    var client = httpClientFactory.CreateClient("AMLClient");
    var response = await client.PostAsync("", new StringContent(requestBody, System.Text.Encoding.UTF8, "application/json"));
    var content = await response.Content.ReadAsStringAsync();
    if (!response.IsSuccessStatusCode)
    {
        return Results.Problem(content, statusCode: (int)response.StatusCode);
    }
    return Results.Json(JsonSerializer.Deserialize<object>(content));
});

app.Run();

Testing the Endpoint

Use curl or Postman to send a request to your web app:

# Example JSON payload
{
    "data": [
        [5.1, 3.5, 1.4, 0.2],
        [6.2, 3.4, 5.4, 2.3]
    ]
}

# Curl example
curl -X POST http://localhost:5000/predict \
     -H "Content-Type: application/json" \
     -d '{"data":[[5.1,3.5,1.4,0.2],[6.2,3.4,5.4,2.3]]}'

Securing Your Web App

Next Steps