NAV Navbar
cURL
  • Introduction
  • Playground
  • Introduction

    Full service API

    Poleon is an AI as a service API that process temporal data of our clients with external data (weather, social events, traffic, ...) to provide real time predictions on where and when citizens will need their services.

    To get an access to our full API that provides real time predictions on transportation demand, please reach us:

    contact@poleon.co

    Playground

    Basic time-series

    curl -X POST \
      https://api.poleon.co/beta1/predict \
      -H 'content-type: multipart/form-data; boundary=----WebKitFormBoundary7MA4YWxkTrZu0gW' \
      -F dataset=@/home/user/dataset.csv \
      -F 'delimiter=;'
      -F date_fieldname=record_date \
      -F value_fieldname=value \
      -F email=result@example.com
    

    The above command returns JSON structured like this:

    {
      "status": "ok or error",
      "message": "optional message",
    }
    

    And you will receive predition results as a mail as soon as they are ready!

    This Playground is a sample you can try:

    Check out the API tester to experiment it even without any technical knowledge.

    You need a CSV file containing 2 columns.

    Unlike our Full Service which will send back real time prediction through the API, our Playground API will send the predicted dataset by email within a few minutes.

    HTTP Request

    POST https://api.poleon.co/beta1/predict

    Query Parameters

    All the query parameters must be passed as multipart/form-data.

    Parameter Mandatory Default Description
    email True Your mail address to receive the results.
    dataset True The dataset you want to predict on, as a CSV.
    delimiter True The delimiter used by the CSV. Ex: , or ;
    value_fieldname False value The name of the field containing the value
    date_fieldname False date The name of the field containing the date
    date_format True The date format (see below)

    date_format follow the Python's strftime format. It is very easy to use. You can get more informations here.

    As an example, %Y-%m will match 2017-09 and 1998-11. But it won't match 10-20-1999.