PayPal Data Scientist interview questions

How do you approach a crypto product

1. Display Number of customers, merchant customers and consumer customers by country sorted by desending order of number of all customers that have atleast 500 customers 2.Display For each Year of transaction, number of transactions, number of transactions with amount >50 USD, number of transactions with amount > 100 USD 3.Assume transaction date gets stored as a string of format mm-dd-yyyy, and one cust_id became stored as null across all tables, and assume leading spaces in cust_id stored in transaction data, How do you find the number of transactions, and transaction value in USD 4. Bin customers into their total transaction value (USD) in 2024, bin the customers into 90-100 percentile , 80 to 90 percentile .... 0-10 percentile ### before you start: there could be some customers who have 0 transaction (these will not be transaction table but they will be in dim_customer table) ### Display ### Percentile bin || Total transaction value (USD) || Total number of customers || Transaction value per customer Sort the number of transactors for each country at a daily level #### Create groups using the following logic, for each country, if the today's number of transactors is + or - 5% of yesterdays transactors keep them in the same group, otherwise assign them to a new group #### Free hint : group can be integer numbers as values. Create a Transactional behaviour table (using common table expressions optionally) and randomly sample 5 records from the above table and print the results