Business Account Data Verification

Build a "Know Your Customer" (KYC) system that reads business accounts from a CSV string and decides whether each one is VERIFIED or NOT VERIFIED. The problem grows in five steps, layering on missing-field checks, description-length limits, blocked generic words, business-name-to-description matching, and finally specific error codes. Problem Requirements You are building a system to verify new business accounts. This is a "Know Your Customer" (KYC) system. The system reads business data from a CSV string. You must decide if each account is VERIFIED or NOT VERIFIED. You will build the solution in five steps. Each step adds a new rule: Check for Missing Info: Make sure all fields have data. Check Description Length: Make sure the description is the right length. Block Bad Words: Flag businesses using forbidden generic words. Match Business Names: Check if the name matches the description fields. Return Specific Error Codes: Tell the user exactly why verification failed. Input Format You get a string formatted as a CSV (Comma Separated Values). col1,col2,col3,col4,col5,col6 BIZ001,Pawsome Pets Inc.,pawsomepets.com,Pawsome,PAWSOME PETS INC,Premium pet supplies BIZ002,Bean Bliss Coffee,beanbliss.com,,,Artisan coffee roasters The first line is the header. Every line after that is a business account with 6 columns: col1: Business ID col2: Legal Name col3: Website URL col4: Short Description col5: Full Description col6: Product Details Output Format For every business (skip the header), print one line like this: VERIFIED: <business_name> Or like this: NOT VERIFIED: <business_name> You must keep the output in the same order as the input. Part 1: Check for Missing Info What You Need To Do Write a function called validate_businesses(csv_data). This function checks if a business provided all the required information. A business is VERIFIED only if every field has text in it. If a field only contains spaces, count it as empty. Example Input: csv_data = """col1,col2,col3,col4,col5,col6 BIZ001,Pawsome Pets Inc.,pawsomepets.com,Pawsome,PAWSOME PETS INC,Premium pet supplies BIZ002,Bean Bliss Coffee,beanbliss.com,,,Artisan coffee roasters BIZ003,,,,, BIZ004,Tech Solutions,techsol.io,Tech,TECH SOLUTIONS,Software consulting""" Output: VERIFIED: Pawsome Pets Inc. NOT VERIFIED: Bean Bliss Coffee NOT VERIFIED: VERIFIED: Tech Solutions Explanation: BIZ001: Has all 6 fields. -> VERIFIED BIZ002: Missing col4 and col5. -> NOT VERIFIED BIZ003: Only has col1. -> NOT VERIFIED (Name is empty, so output name is blank) BIZ004: Has all fields. -> VERIFIED Rules All 6 columns must have data (remove whitespace first). Ignore the first row (the header). Keep the order of the businesses the same. Handle cases where the business name (col2) is missing. Questions to Ask the Interviewer Can I assume the CSV format is perfect, or should I check for errors? What if a row is missing columns completely? What do I print if the business name itself is empty? Do special characters count as "empty"? Part 2: Check Description Length What You Need To Do Update your code to check the length of the Full Description (col5). It must follow Stripe's rules. The length must be between 5 and 31 characters (inclusive). Remove extra spaces before counting. If the length is wrong, mark it NOT VERIFIED. Example Input: csv_data = """col1,col2,col3,col4,col5,col6 BIZ001,Pawsome Pets Inc.,pawsomepets.com,Pawsome,PAWSOME PETS INC,Premium pet supplies BIZ002,Bean Bliss Coffee,beanbliss.com,Bean,Bean,Artisan coffee roasters BIZ003,Oakridge Furniture,oakridge.com,Oak,OAKRIDGE CUSTOM WOODWORKING AND FURNITURE EMPORIUM,Custom furniture BIZ004,Tech Solutions,techsol.io,Tech,ITCS,Software consulting""" Output: VERIFIED: Pawsome Pets Inc. NOT VERIFIED: Bean Bliss Coffee NOT VERIFIED: Oakridge Furniture NOT VERIFIED: Tech Solutions Explanation: BIZ001: Length is 16. This is okay. -> VERIFIED BIZ002: Length is 4. Too short. -> NOT VERIFIED BIZ003: Length is 50. Too long. -> NOT VERIFIED BIZ004: Length is 4. Too short. -> NOT VERIFIED Rules Remove whitespace from col5 before checking. Length must be between 5 and 31. Check this after checking for missing fields (Part 1). Part 3: Block Bad Words What You Need To Do To stop fraud, we must block generic business names in the Full Description (col5). If col5 contains any of these words, mark the account NOT VERIFIED: Blocked Words (Case-Insensitive): ONLINE STORE ECOMMERCE RETAIL SHOP GENERAL MERCHANDISE Case-insensitive means "Shop", "SHOP", and "shop" are all blocked. Example Input: csv_data = """col1,col2,col3,col4,col5,col6 BIZ001,Pawsome Pets Inc.,pawsomepets.com,Pawsome,PAWSOME PETS INC,Premium pet supplies BIZ002,Global Goods Market,globalgoods.com,Global,ONLINE STORE,Various products BIZ003,Northwest Tech,nwtech.com,NW Tech,NORTHWEST INNOVATION TECH,Technology solutions BIZ004,Sweet Dreams,sweetdreams.com,Sweet,SWEET DREAMS CREAMERY,Ice cream shop""" Output: VERIFIED: Pawsome Pets Inc. NOT VERIFIED: Global Goods Market VERIFIED: Northwest Tech VERIFIED: Sweet Dreams Explanation: BIZ001: No bad words. -> VERIFIED BIZ002: Contains "ONLINE STORE". -> NOT VERIFIED BIZ003: No bad words. -> VERIFIED BIZ004: col5 does not have bad words (we ignore col6). -> VERIFIED Rules Check col5 for blocked terms. Ignore upper/lower case. Do this check in addition to Part 1 and Part 2. Questions to Ask the Interviewer Do I block the word if it is part of another word? (e.g., does "WORKSHOP" count as "SHOP"?) Are there other words to block? Should I check other columns too? Part 4: Match Business Names What You Need To Do Make sure the Business Name (col2) matches the descriptions (col4 or col5). At least 50% of the words in the Business Name must appear in either the Short Description (col4) or Full Description (col5). How to Match Words: Split col2, col4, and col5 into words using spaces. Remove "LLC" and "Inc" from the lists. Ignore upper/lower case. Check if half of the name's words exist in the description fields. Example Input: csv_data = """col1,col2,col3,col4,col5,col6 BIZ001,land water,landwater.com,land,land water LLC,Environmental services BIZ002,Acme Global Trading,acme.com,Acme,XYZ ENTERPRISES,Import export services BIZ003,Maple Ridge Bakery,maplebakery.com,Maple,MAPLE RIDGE BAKERY LLC,Artisan baked goods BIZ004,Innovation Labs Inc,innovlabs.com,Labs,INNOVATION RESEARCH,R&D services""" Output: VERIFIED: land water NOT VERIFIED: Acme Global Trading VERIFIED: Maple Ridge Bakery VERIFIED: Innovation Labs Inc Explanation: BIZ001: "land water" (2 words). Both are in col5. Match: 100%. -> VERIFIED BIZ002: "Acme Global Trading" (3 words). Only "Acme" is found. Match: 33% (less than 50%). -> NOT VERIFIED BIZ003: "Maple Ridge Bakery" (3 words). All 3 match. Match: 100%. -> VERIFIED BIZ004: "Innovation Labs Inc" (2 words, ignoring Inc). Both match. Match: 100%. -> VERIFIED Rules Split text by whitespace. Ignore "LLC" and "Inc". Comparison is case-insensitive. You need >= 50% match. Questions to Ask the Interviewer Can I combine col4 and col5 for the check? How do I handle punctuation (like "Ben & Jerry's")? What if the name becomes empty after removing "LLC"? Part 5: Return Specific Error Codes What You Need To Do Do not just say "NOT VERIFIED". Return a specific code explaining the error. If there are multiple errors, return the first one you find. Order of Checks: Empty fields → ERROR_MISSING_FIELDS Bad length → ERROR_INVALID_LENGTH Blocked word → ERROR_GENERIC_NAME Name mismatch → ERROR_NAME_MISMATCH Example Input: csv_data = """col1,col2,col3,col4,col5,col6 BIZ001,Pawsome Pets Inc.,pawsomepets.com,Pawsome,PAWSOME PETS INC,Premium pet supplies BIZ002,Bean Bliss Coffee,beanbliss.com,,,Artisan coffee roasters BIZ003,Short Name,short.com,Short,SHRT,Products BIZ004,Generic Store,generic.com,Store,RETAIL,Various items BIZ005,Mismatched Corp,mismatch.com,Wrong,DIFFERENT BUSINESS,Services""" Output: VERIFIED: Pawsome Pets Inc. ERROR_MISSING_FIELDS: Bean Bliss Coffee ERROR_INVALID_LENGTH: Short Name ERROR_GENERIC_NAME: Generic Store ERROR_NAME_MISMATCH: Mismatched Corp Rules Check rules in the order listed above. Stop at the first error. Only return VERIFIED if everything passes. Solution Details Step 1: Checking for Missing Info Plan: Read the CSV string and split it by lines. Ignore the first line (header). Loop through every data row: Split the row by commas. Check if all 6 items have text. Print the result. Complexity: O(n × m) (n = rows, m = row length). Memory: O(n) to store the lines. Code: def validate_businesses(csv_data): lines = csv_data.strip().split('\n') results = [] for i, line in enumerate(lines): if i == 0: # Skip header continue fields = [field.strip() for field in line.split(',')] # Ensure we have exactly 6 fields if len(fields) != 6: results.append(f"NOT VERIFIED: {fields[1] if len(fields) > 1 else ''}") continue # Check if all fields are non-empty all_filled = all(field for field in fields) business_name = fields[1] if all_filled: results.append(f"VERIFIED: {business_name}") else: results.append(f"NOT VERIFIED: {business_name}") return '\n'.join(results) Things to Watch Out For: Empty lines. Rows with missing columns. Fields with only spaces. Step 2: Checking Description Length Plan: Add a check for column 5 length. Do this after ensuring fields are not empty. Range is 5 to 31. Code: def validate_businesses(csv_data): lines = csv_data.strip().split('\n') results = [] for i, line in enumerate(lines): if i == 0: # Skip header continue fields = [field.strip() for field in line.split(',')] if len(fields) != 6: results.append(f"NOT VERIFIED: {fields[1] if len(fields) > 1 else ''}") continue business_name = fields[1] full_descriptor = fields[4] # Check 1: All fields non-empty if not all(field for field in fields): results.append(f"NOT VERIFIED: {business_name}") continue # Check 2: Descriptor length descriptor_len = len(full_descriptor) if descriptor_len < 5 or descriptor_len > 31: results.append(f"NOT VERIFIED: {business_name}") continue results.append(f"VERIFIED: {business_name}") return '\n'.join(results) Note: We use strip() to remove spaces before checking length. Step 3: Blocking Generic Names Plan: Create a list of blocked words. Check if col5 contains any of them. Convert text to uppercase so case doesn't matter. Code: def validate_businesses(csv_data): BLOCKED_TERMS = { 'ONLINE STORE', 'ECOMMERCE', 'RETAIL', 'SHOP', 'GENERAL MERCHANDISE' } lines = csv_data.strip().split('\n') results = [] for i, line in enumerate(lines): if i == 0: continue fields = [field.strip() for field in line.split(',')] if len(fields) != 6: results.append(f"NOT VERIFIED: {fields[1] if len(fields) > 1 else ''}") continue business_name = fields[1] full_descriptor = fields[4] # Check 1: All fields non-empty if not all(field for field in fields): results.append(f"NOT VERIFIED: {business_name}") continue # Check 2: Descriptor length if len(full_descriptor) < 5 or len(full_descriptor) > 31: results.append(f"NOT VERIFIED: {business_name}") continue # Check 3: Blocked terms descriptor_upper = full_descriptor.upper() has_blocked_term = any(term in descriptor_upper for term in BLOCKED_TERMS) if has_blocked_term: results.append(f"NOT VERIFIED: {business_name}") continue results.append(f"VERIFIED: {business_name}") return '\n'.join(results) Note: This logic checks if the blocked term is inside the text (substring match). Step 4: Matching Business Names Plan: Split names into word lists. Filter out "LLC" and "Inc". Count how many name words appear in the descriptions. If matches / total_words >= 0.5, it passes. Code: def validate_businesses(csv_data): BLOCKED_TERMS = { 'ONLINE STORE', 'ECOMMERCE', 'RETAIL', 'SHOP', 'GENERAL MERCHANDISE' } IGNORED_WORDS = {'llc', 'inc'} def get_words(text): """Extract and normalize words, removing ignored terms.""" words = text.lower().split() return [w for w in words if w not in IGNORED_WORDS] def check_name_match(business_name, short_desc, full_desc): """Check if at least 50% of business name words match descriptors.""" name_words = set(get_words(business_name)) short_words = set(get_words(short_desc)) full_words = set(get_words(full_desc)) if not name_words: # Edge case: no valid words in business name return False # Count matches in either short or full descriptor combined_descriptor_words = short_words | full_words matches = len(name_words & combined_descriptor_words) match_percentage = matches / len(name_words) return match_percentage >= 0.5 lines = csv_data.strip().split('\n') results = [] for i, line in enumerate(lines): if i == 0: continue fields = [field.strip() for field in line.split(',')] if len(fields) != 6: results.append(f"NOT VERIFIED: {fields[1] if len(fields) > 1 else ''}") continue business_name = fields[1] short_descriptor = fields[3] full_descriptor = fields[4] # Check 1: All fields non-empty if not all(field for field in fields): results.append(f"NOT VERIFIED: {business_name}") continue # Check 2: Descriptor length if len(full_descriptor) < 5 or len(full_descriptor) > 31: results.append(f"NOT VERIFIED: {business_name}") continue # Check 3: Blocked terms descriptor_upper = full_descriptor.upper() if any(term in descriptor_upper for term in BLOCKED_TERMS): results.append(f"NOT VERIFIED: {business_name}") continue # Check 4: Name consistency if not check_name_match(business_name, short_descriptor, full_descriptor): results.append(f"NOT VERIFIED: {business_name}") continue results.append(f"VERIFIED: {business_name}") return '\n'.join(results) Key Details: Uses set for faster lookups. & finds the intersection (common words). | joins two sets together. Complexity: O(n × w) (w = average words per field). Step 5: Returning Error Codes Plan: Keep the order of checks the same. Change the output string to the specific error code. Return immediately when an error is found. Code: def validate_businesses(csv_data): BLOCKED_TERMS = { 'ONLINE STORE', 'ECOMMERCE', 'RETAIL', 'SHOP', 'GENERAL MERCHANDISE' } IGNORED_WORDS = {'llc', 'inc'} def get_words(text): words = text.lower().split() return [w for w in words if w not in IGNORED_WORDS] def check_name_match(business_name, short_desc, full_desc): name_words = set(get_words(business_name)) short_words = set(get_words(short_desc)) full_words = set(get_words(full_desc)) if not name_words: return False combined_descriptor_words = short_words | full_words matches = len(name_words & combined_descriptor_words) return matches / len(name_words) >= 0.5 lines = csv_data.strip().split('\n') results = [] for i, line in enumerate(lines): if i == 0: continue fields = [field.strip() for field in line.split(',')] if len(fields) != 6: results.append(f"ERROR_MISSING_FIELDS: {fields[1] if len(fields) > 1 else ''}") continue business_name = fields[1] short_descriptor = fields[3] full_descriptor = fields[4] # Validation 1: All fields non-empty if not all(field for field in fields): results.append(f"ERROR_MISSING_FIELDS: {business_name}") continue # Validation 2: Descriptor length if len(full_descriptor) < 5 or len(full_descriptor) > 31: results.append(f"ERROR_INVALID_LENGTH: {business_name}") continue # Validation 3: Blocked terms descriptor_upper = full_descriptor.upper() if any(term in descriptor_upper for term in BLOCKED_TERMS): results.append(f"ERROR_GENERIC_NAME: {business_name}") continue # Validation 4: Name consistency if not check_name_match(business_name, short_descriptor, full_descriptor): results.append(f"ERROR_NAME_MISMATCH: {business_name}") continue results.append(f"VERIFIED: {business_name}") return '\n'.join(results) Extra Questions to Prepare For Question: What tricky edge cases might break this code? CSV Issues: Commas inside fields: If a name is "Smith, Jones, and Co", simple splitting fails. You need a real CSV parser. Special Characters: Emojis or accents in names. Huge Files: If the file is too big for memory, you must process it line-by-line using a stream. Code Example for Robustness: import csv from io import StringIO def validate_businesses_robust(csv_data): """Handle edge cases like fields containing commas.""" reader = csv.reader(StringIO(csv_data)) next(reader) # Skip header results = [] for row in reader: if len(row) != 6: # Handle inconsistent column count continue # Process with proper CSV escaping Data Issues: Weird Names: A business named "Inc" would have 0 words after filtering. Abbreviations: "IBM" vs "International Business Machines". Security: Users trying to inject SQL or HTML into the fields.