curl --request POST \
  --url https://api.orbit.devotel.io/api/v1/cdp/predictive-models/{model}/score \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "limit": 25
}'

{
  "data": {
    "model": "<string>",
    "kind": "classification",
    "output_unit": "<string>",
    "population_scored": 123,
    "limit": 123,
    "count": 123,
    "scored": {
      "contact_id": "<string>",
      "display_name": "<string>",
      "score": 123,
      "predicted_value_cents": 123
    }
  },
  "meta": {
    "request_id": "<string>",
    "timestamp": "2023-11-07T05:31:56Z",
    "docs_url": "<string>"
  }
}

CDP

Score profiles with a predictive model

Score profiles with the named model, reusing the cached fit from the most recent train (fitting on demand if none is cached). Post a contact_id to score exactly one contact, or omit it to score the model’s activation population and return the top-ranked profiles (highest score/value first) up to limit. Classification models return a probability score; the lifetime-value model returns predicted_value_cents. Returns 422 when the model cannot be fit, and 404 when a requested contact_id is not in the scoring population. Owner / admin / developer only.

POST

api

cdp

predictive-models

{model}

score

curl --request POST \
  --url https://api.orbit.devotel.io/api/v1/cdp/predictive-models/{model}/score \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "limit": 25
}'

{
  "data": {
    "model": "<string>",
    "kind": "classification",
    "output_unit": "<string>",
    "population_scored": 123,
    "limit": 123,
    "count": 123,
    "scored": {
      "contact_id": "<string>",
      "display_name": "<string>",
      "score": 123,
      "predicted_value_cents": 123
    }
  },
  "meta": {
    "request_id": "<string>",
    "timestamp": "2023-11-07T05:31:56Z",
    "docs_url": "<string>"
  }
}

Authorizations

Authorization

string

header

required

Dashboard JWT token from Clerk

Path Parameters

model

enum<string>

required

Model key from GET /api/v1/cdp/predictive-models.

Available options:

churn_propensity,

conversion_intent,

lifetime_value,

engagement_fatigue

Body

application/json

contact_id

string

Score just this contact. Omit to score and rank the whole activation population.

Required string length: 1 - 128

limit

integer

default:50

Maximum number of ranked profiles to return in population mode.

Required range: 1 <= x <= 500

Response

Scored profile(s) for the model.

data

object

Show child attributes

API Reference

All Endpoints

Score profiles with a predictive model

Authorizations

Path Parameters

Body

Response