How do I practice coding and algorithm questions?

Use PracHub's coding console to write, test, and debug your solutions in Python or JavaScript. View hints, test against sample inputs, and compare with official solutions.

What difficulty level is this coding question?

This is a medium difficulty Coding & Algorithms question, commonly asked during Technical Screen rounds at Stripe.

What role is this question designed for?

This question is commonly asked for Software Engineer candidates at Stripe during technical interviews.

Validate KYC CSV Records | Stripe Coding Question

Validate KYC CSV Records

Company: Stripe

Role: Software Engineer

Category: Coding & Algorithms

Difficulty: medium

Interview Round: Technical Screen

Implement a function that validates business-verification records from a CSV-formatted string. - The first line is a header with exactly 6 comma-separated columns: `col1,col2,col3,col4,col5,col6`. - Each remaining line is a data row with exactly 6 comma-separated fields. - You may assume fields do not contain embedded commas or quoted newlines. For each data row, return `VERIFIED` if all of the following rules hold; otherwise return `NOT VERIFIED`. 1. **No empty fields** - All 6 fields must be present. - After trimming leading and trailing whitespace, every field must be non-empty. 2. **Column 5 length constraint** - After trimming whitespace, the length of `col5` must be between `5` and `31`, inclusive. 3. **Forbidden terms in `col2`** - Case-insensitively, `col2` must not contain any of these terms as substrings: `company`, `firm`, `co.`, `corporation`, `group`. 4. **Word overlap between `col2` and `col4` or `col5`** - Convert `col2`, `col4`, and `col5` to lowercase and split each one by whitespace into words. - Ignore the words `llc` and `inc` in all three columns. - Let `W2` be the remaining words from `col2`. - The row passes this rule if at least 50% of the words in `W2` also appear in the words from `col4`, or at least 50% of the words in `W2` also appear in the words from `col5`. - Matching is exact string equality after lowercasing. - You may assume `W2` is non-empty after removing ignored words. Return the verification result for each data row in the original order.

Quick Answer: This question evaluates skills in CSV parsing, string normalization, field-level validation, and rule-based text matching, emphasizing attention to edge cases such as trimming whitespace, substring checks, and word-overlap logic.

Implement a function that validates business-verification records from a CSV-formatted string. The input always starts with the header `col1,col2,col3,col4,col5,col6`, followed by zero or more data rows. Return one result for each data row in order: `VERIFIED` if the row satisfies every rule below, otherwise `NOT VERIFIED`. Validation rules for each data row: 1. The row must contain exactly 6 comma-separated fields. 2. After trimming leading and trailing whitespace, all 6 fields must be non-empty. 3. After trimming whitespace, the length of `col5` must be between 5 and 31 inclusive. 4. Case-insensitively, `col2` must not contain any of these forbidden substrings: `company`, `firm`, `co.`, `corporation`, `group`. 5. Let `W2` be the words in `col2` after lowercasing, splitting by whitespace, and removing the words `llc` and `inc`. Do the same for `col4` and `col5`. - The row passes this rule if at least 50% of the words in `W2` also appear in the words from `col4`, or at least 50% of the words in `W2` also appear in the words from `col5`. - Word matching is exact string equality after lowercasing. - You may assume `W2` is non-empty for valid test data.

Constraints

1 <= total length of `csv_data` <= 10^6
The first line is the header `col1,col2,col3,col4,col5,col6`
Fields do not contain embedded commas or quoted newlines
For overlap checking, split words only by whitespace
`W2` can be assumed non-empty after removing `llc` and `inc`

Examples

Input: ("col1,col2,col3,col4,col5,col6",)

Expected Output: []

Explanation: There are no data rows, so the result is an empty list.

Input: ("col1,col2,col3,col4,col5,col6\n1,Blue Ocean LLC,x,Ocean Blue,Blue LLC,z\n2,North GROUP Holdings,x,North Holdings,North Holdings,z\n3,Sunrise Inc Bakery,x,Random Name, Sunrise Bakery ,z\n4,Red Apple Bakery Cafe,x,Apple Cafe,ValidFive,z\n5,Green Field Market,x,Green Shop,Field,z",)

Expected Output: ["VERIFIED", "NOT VERIFIED", "VERIFIED", "VERIFIED", "NOT VERIFIED"]

Explanation: Row 1 matches fully with col4. Row 2 contains the forbidden term 'group'. Row 3 matches fully with col5 after trimming and ignoring 'inc'. Row 4 has exactly 50% overlap with col4 (2 of 4 words). Row 5 has only 1 of 3 words overlapping with either col4 or col5.

Input: ("col1,col2,col3,col4,col5,col6\n1,Alpha Beta,x,Alpha Beta,abcde,z\n2,Alpha Beta,x,Alpha Beta, ,z\n3,Alpha Beta,x,Alpha,abcd,z",)

Expected Output: ["VERIFIED", "NOT VERIFIED", "NOT VERIFIED"]

Explanation: Row 1 is valid and uses the lower bound length 5 for col5. Row 2 fails because col5 is empty after trimming. Row 3 fails because col5 has length 4.

Input: ("col1,col2,col3,col4,col5,col6\n1,North Star LLC,x,North Star,1234567890123456789012345678901,z\n2,Fresh Market,x,Fresh,12345",)

Expected Output: ["VERIFIED", "NOT VERIFIED"]

Explanation: Row 1 is valid and uses the upper bound length 31 for col5. Row 2 has only 5 fields instead of 6, so it is not verified.

Hints

Process one row at a time: split by commas, trim each field, and fail fast as soon as any rule is broken.
For the 50% overlap rule, keep `col2` as a list of words, but convert the words from `col4` and `col5` to sets for fast membership checks. Use `2 * matches >= len(W2)` to test the threshold without floating-point math.

Quick Overview

This question evaluates skills in CSV parsing, string normalization, field-level validation, and rule-based text matching, emphasizing attention to edge cases such as trimming whitespace, substring checks, and word-overlap logic.