Skip to content

eriks-v/premiums

Repository files navigation

Implementation notes

PremiumCalculator implements interface CalculatePremium. Application is split in two parts Premium calculation and Policy Validation

Application specific entities

Premium

Contains premium amount and implements premium amount validation logic.

PremiumAttribute

Contains aggregated data for premium calculation logic (Which coefficient to use).

Policy Pricing

Policy pricing is split int two parts. PremiumAttribte aggregation and pricing.

PremiumAttribute aggregation

PremiumAttributeService is used to gather all PolicySubObjects and do aggregation. Result is collection of PremiumAttributes that is used for Premium calculation. In case when PremiumAttribute collection is empty* exception is thrown.

As per definition, there can exist PolicyObjects without assigned PolicySubObjects, which should be treated (according to answers provided) as if nothing is insured. In case when policy contains only PolicyObyest that do not have any PolicySubObjects we have "empty" policy and no meaningful calculation is possible. In this case exception is thrown.

Pricing

PremiumAttribute is used to callculate premium amount in PolicyPricingService Pricing logic is implemented in package lv/pi/premiums/application/service/pricing. Pricing is done by choosing the correct pricing coefficient provided by the matching pricing rule. The pricing rule implements the interface PremiumPricingRule and any decision interfaces that contain default methods for checking for specific risk types like isFire and sumInsured amount range. Pricing rules are marked with annotations @Component and @Order as rule evaluation order matters.

PolicyPricingService handles pricing rules. It contains a collection of orderd pricing rules. Spring injects pricing rules in collection which imlpement PremiumPricingRule interface. PricingAttributes are passed to a service that evaluates rules one by one and uses the first matching rule to retrieve the coefficient for specific PolicyAttribute. A coefficient is used to calculate an amount for specific PolicyAttribute. If multiple PolicyAttributes were provided to PolicyPricingService calculated amounts are summed. The total amount is returned to PremiumCalculator where FinancialRoundingService is used to round the value to configured scale.

If no matching rule is found, PremiumCalculationException is thrown.

Pricing implementation diagram

class diagram

Policy Validation

As the application does not control Policy creation, it should do Policy validation.

Policy basic validation includes required fields and is done by Lombok @NonNull annotation. Also, the assumption is made that object collections all are Set's to enforce the uniqueness of entries. Fields "Risk type" and "Policy status" value (e.g., FIRE, THEFT) enforcement was done by implementing those fields as enums.

Advanced validation logic is implemented in package lv/pi/premiums/application/service/validation. Idea behind implementation is that it is possible to define a validation rule that can be implemented on any entity.

Validation rule implements ValidationRule parametrized interface that defines method validate that implements validation logic on entity. Rule implementation is annotated with spring annotation @Component. @Order annotation should be used if rule evaluation order matters. In the current implementation, @Order is used to ensure null validation happens first. PolicyValidationException can be used if the validation does not pass.

Entity-specific validation services are created that extend parametrized abstract class Validate that contains a collection of entity-specific rules. Validation rule collection is populated by spring injecting components that implement interface ValidationRule. All specific entity rules are evaluated during validation.

Validation implementation diagram

class diagram

Application specific entity validation

As application controls creation of entities Premium and PremiumAttribute, validation is done on entity creation.

Adding new risk type

  1. Add a new constant to RiskType with a new risk type name.
  2. Create a new pricing rule class that implements the PremiumPricingRule interface.
    1. It is preferred to create a new interface that will implement the default method for identifying new risk type. e.g., FirePricingDecision.

Example run in PremiumsApplication

@SpringBootApplication
@Slf4j
public class PremiumsApplication {

    public static void main(String[] args) {
        ApplicationContext applicationContext = SpringApplication.run(PremiumsApplication.class, args);

        CalculatePremium premiumCalculator = applicationContext.getBean(CalculatePremium.class);

        Policy policy =
                new Policy("LV20-02-100000-5", PolicyStatus.REGISTERED, Set.of(
                        new PolicyObject("po-1",
                                Set.of(new PolicySubObject("pso-1", new BigDecimal("500"), RiskType.FIRE),
                                        new PolicySubObject("pso-2", new BigDecimal("102.51"), RiskType.THEFT)
                                )
                        )));

        log.info("Rez: " + premiumCalculator.calculate(policy).getAmount());
    }
}

Prerequisites

  • JDK 11
  • Project uses Lombok, so enable annotation processing in your IDE

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages