Cuvva take home challenge

Thanks for applying to become a Cuvva Backend Engineer 🎉. Below is a challenge we'd like you complete and send back to us.

The task

Cuvva sells a lot of motor insurance, you can purchase cover for a range of durations, from 30 minutes all the way up to 28 days. To do this we need a system that prices a user accurately and quickly.

Pricing means taking a set of information (usually details about a user) and ultimately generating a single price. This is the 'quote price' that we will ultimately end up charging a customer.

In this scenario, the Product Manager has asked you to come up with a system to perform pricing calculations on a single user for a range of durations. The PM mentioned that the system needs to potentially be able to reprice thousands of users within a single minute. They also mentioned that the price needs to be deterministic, given the same input information it should always give you the same output.

Resources

You will find attached

• A code stub in Go implementing an RPC endpoint
• A model, this is an excel sheet detailing how we translate risks about the user to a price. Link here.

The Challenge

Task 1 - Familiarise yourself with the resources

Take a look at the code stub and familiarise yourself with the structure. This implements an RPC endpoint (generate_pricing) which is called with a single users details.

Take a look at the provided model. The model gives you details on how to take in a risk i.e. an insurance group and translate it to a "factor". These factors get multiplied with the "Base Rate" price (found in the "Base Rates" sheet) to form a total price the user should pay. The base rate changes depending on the duration of cover selected. We typically give the user a price for every duration of cover we offer, letting the user pick the duration that makes sense for them. Take a look at the "Calculator" sheet and familiarise yourself with how the total price a user pays is calculated.

Task 2 - Accept details about a user

Update the generate_pricing endpoint to accept the following example input.

{
    "date_of_birth": "1970-12-04",
    "insurance_group": 12,
    "license_held_since": "1988-08-01"
}

This represents details about a single user.

Task 3 - Produce a users price for a range of durations

Update the generate_pricing endpoint to respond with the total price a user should pay for each duration of cover found in the base rates sheet.

The output must be in JSON, the schema of the output is left to you.

Rules & time limits

Below are some rules on how much you should feel you need to do:

• The recommended time to complete the challenge is four hours
  • If you reach the recommended time and haven't completed the task in the way you want, you can use a README to document what you would have liked to do. If you do so, please be as detailed as possible.
  • Or if you do want to put more time into the task, feel free to do so and mention this when you submit the solution.
• You should write code as if it were going to go into production. This means using best practices where possible and having an appropriate code structure.
  • We favour well-documented, simple code over code that is complex, terse and hard to read.
  • We favour well tested code.
  • We favour making proper use of concurrency to speed up processes where possible.
  • We believe it is good practice to make sure inputs are validated and sanitized before being used.
  • We expect you to have an appropriate code design and layout which would allow the application to scale to a large number of pricing factors (30-50), while remaining easy to read and maintain for developers.
• We do not intend you to read and parse the Excel spreadsheet we provided. This is only provided to communicate the model. Feel free to copy the factors into a structure you find appropriate for the exercise.
• Once you're finished we'd like you to submit your code back to us as a zip. Please do not use a public repository.