🚀 Get to pre-production in weeks, not months, with private training direct from Jube’s developer — real sovereignty, zero vendor lock-in.

HTTP Adaptation

HTTP Adaptation refers to the dispatch of the full transaction payload in a POST body to a remote HTTP endpoint for the receipt of a quantitative score, and optionally a structured explanation of it, in the JSON response body.

The request body is the entire EntityAnalysisModelInstanceEntryPayload for the transaction - Payload, Abstraction, TTL Counters, Abstraction Calculations, Sanctions, and any HTTP Adaptations with a lower Priority that have already been recalled for the same transaction.

The response body may still be a bare number for the simplest case:

0.91

Alternatively, a structured object carrying a score alongside model metadata, calibration, contribution and journey - see the HTTP Adaptation Protocol page for the full response specification, consumption rules, and a set of mock reference endpoints to test against.

The intention of HTTP Adaptation is to recall R models via Plumber, Python models via Flask or make use of any HTTP service that respects the payload specification set out above.

To create a HTTP Adaptation, navigate Models » Machine Learning » HTTP Adaptation:

Click on a model in the tree to create a new HTTP Adaptation:

The HTTP Adaptation accepts the following parameters:

Value	Description	Example
HTTP Endpoint	Using the prefix specified in the HttpAdaptationUrl Environment Variable in concatenation with remainder of the URL for the HTTP Endpoint to POST to. Assume no “/” terminating the HttpAdaptationUrl Environment Variable value.	/api/invoke/ExampleFraudScoreLocalEndpoint
Priority	Ascending execution order among the adaptations on a model, to support boosting/model chaining - a lower Priority adaptation is recalled first, and its result is available to a later one via the request body. Defaults to 0.	0.1

In this example an endpoint is available for the purpose of echoing back the Square Root of the ResponseCodeEqual0Volume Abstraction Rule Value at https://localhost:5001/api/invoke/ExampleFraudScoreLocalEndpoint. Complete the page as follows:

Scroll down and click Add to create a version of the HTTP Adaptation:

Synchronise the model via Entity » Synchronisation and repeat the HTTP POST to endpoint https://localhost:5001/api/invoke/EntityAnalysisModel/90c425fd-101a-420b-91d1-cb7a24a969cc for response as follows.

Notice that the score has been returned for use in Activation Rules in the Adaptation entity.