Click the button below to quickly clone and deploy into your own Heroku acount. If you don't have one it'll prompt you to setup a free one.

Once deployed to your Heroku instance run the following:

curl -s -XPOST 'https://<name-of-your-heroku-app>.herokuapp.com/' -d '{"Pclass":3,"Age":2,"SibSp":1,"Fare":50}' -H 'accept-content: application/json'

Alternatively a simple python script:

import requests
import json
url = 'https://<name-of-your-heroku-app>.herokuapp.com/'
data = {"Pclass":3, "Age":2, "SibSp":1, "Fare":50}
response = requests.post(url, json.dumps(data))
print(response.json())

You can read the full article that was published on Towards Data Science

• Article: Towards Data Science: Create an API to Deploy Machine Learning Models using Flask
• Author: Elizabeth Ter Sahakyan

Thank you to Josh Peak who added the button for heroku deployment