People Like Me

This walkthrough demonstrates how a client can get a personalized goal recommendation using only a few personal information with People Like Me (PLM) API.

First the information that the client needs to input is age.

Next the client needs to input monthly income and monthly savings percentage.

The information from the previous steps is enough to get goal recommendation data from the People Like Me API. Here is the API to get goal recommendations.

POST /api/autoMl/v2/prediction

Firstly, it is required to specify clientId and model in the API. In production we will give these values to you after we build an AI model for you using the data that we have or you provided. clientId will be tied to your API key and used to authenticate the usage of the model. model refers to the name of the method used to build the model.

In the example payload below, the clientId and model are sandbox ID and model which everyone can use. To get a production ready clientId and model using data provided by you or us, contact us here.

Next, we will add the customer data gathered from the previous page in the API request payload to do goal recommendation.

{
  "clientId": "usa_prod",
  "model": "ensemble",
  "customer": [
    {
        "nationality": "American",
        "gender": "male",
        "age": 28,
        "race": "American",
        "personalIncome": 8000,
        "savingsRatio": 0.75,
        "platformCountry": "US",
        "customerId": "0"
    }
  ]
}

The response payload will return the goal recommendation with respective probability for every customer. A goal with highest probability will be labelled as the recommended goal for a particular customer.

[
  {
    "customerId": "0",
    "goalRecommendation": [
      {
          "goalType": "retirement",
          "predictedProb": 0.27947
      },
      {
          "goalType": "growWealth",
          "predictedProb": 0.10261
      },
      {
          "goalType": "car",
          "predictedProb": 0.14674
      },
      {
          "goalType": "wedding",
          "predictedProb": 0.79104
      },
      {
          "goalType": "education",
          "predictedProb": 0.49972
      },
      {
          "goalType": "house",
          "predictedProb": 0.65447
      },
      {
          "goalType": "holiday",
          "predictedProb": 0.14989
      },
      {
          "goalType": "financialSecurity",
          "predictedProb": 0.03225
      },
      {
          "goalType": "business",
          "predictedProb": 0.5553
      }
    ]
  }
]

We can then display the goal recommendations as shown below.

There is an information icon at the top right corner of the image above and client is able to find out more details about the goal recommendation and the probability upon clicking on the icon.