How do you import a machine learning model that you trained in python - Decisimo - Decision Intelligence Services

Published on: 2024-08-10 18:48:28

The five basic steps when creating an API:
1.

Design

  • You want to know what decision making process you want to solve using Decisimo. Based on that you need to prepare data for sending to the decision engine. Rule of thumb is to send all data you have related to a situation about which you want to make a decision.
2.

Selecting instance region

  • When selecting an endpoint region for API endpoint, decide based on data privacy and protection rules. Instance is a place where your API endpoint will be established.
    You can create as many endpoints as you wish in various regions.
3.

Testing endpoint

  • Since you can create and have as many endpoints as you wish, a good guide is to create separate endpoints for testing and production. Everytime before deploying your new decision flow to the production API endpoint, it is a good practice to make a test on a testing API endpoint using your favorite testing tool (Postman or SoapUI).
4.

Request and Response payload

  • Decisimo API endpoints consume JSON as input requests sent using method POST. Response from the endpoint is also JSON. The same JSON enriched or changed within decision flow.
5.

Deploy your decision flow

  • Build your first decision flow and deploy it to your endpoint to start running automated decision making.