Prompt You are a helpful and accurate AI assistant. Analyze this receipt text that was extracted and structure the following information in JSON format. Be as accurate as possible and only include information that is clearly present in the text. Product Matching Rules: Product Matching Instructions: Analyze each receipt item and use your knowledge of P&G brands and products to determine if there's a reasonable match. Use your best judgment based on: Known P&G brand names (Pampers, Oral-B, Gillette, Venus, Pantene, Head & Shoulders, Olay, Braun, Ariel, Febreze, Always, Aussie, Clearblue, Pepto-Bismol, etc.) Product type/category similarities Partial product names or model numbers Common abbreviations or variations Make intelligent matching decisions considering that receipt names may be: Truncated due to receipt width limitations Abbreviated or shortened Missing spaces or punctuation Using different terminology for the same product type Output Format: Respond only with the JSON object with the following structure: { "receipt_contents": { "store_name": "Name of the store/business", "store_address": "Complete address if visible", "phone_number": "Phone number if visible", "date": "Date of transaction (MM/DD/YYYY or DD/MM/YYYY format)", "time": "Time of transaction", "receipt_number": "Receipt or transaction number", "items": [ { "original_name": "Item name as it appears on receipt", "matched_name": "Full P&G product name if matched, otherwise same as original_name", "price": "Item price", "quantity": "Quantity if specified. Fallback 1", "category": "Product category (see categories below)", "brand": "Brand name if identified", "match_confidence": "exact|partial|brand_only|none" } ], "subtotal": "Subtotal amount", "tax": "Tax amount", "total": "Total amount", "payment_method": "Payment method used (cash, card, etc.)", "cashier": "Cashier name if visible", "additional_info": "Any other relevant information like discounts, promotions, loyalty points, etc." }, "matched_products": [ { "original_name": "Name as it appeared on receipt", "matched_name": "Full matched product name", "total_price": "Item total price (price * quantity)", "category": "Product category", "brand": "Brand name", "match_confidence": "exact|partial|brand_only|none" } ], "summary": { "total_items": "Number of items", "matched_items": "Number of successfully matched P&G products", "unmatched_items": "Number of items that couldn't be matched to P&G products" } } Product Categories: Personal Care - Oral Care: Oral-B products (toothbrushes, toothpaste, etc.) Personal Care - Grooming: Braun products (epilators, shavers, etc.) Personal Care - Shaving: Gillette, Venus products (razors, shaving cream, etc.) Personal Care - Skincare: Olay products (serums, moisturizers, etc.) Hair Care: Pantene, Aussie, Head & Shoulders products Baby Care: Pampers products (diapers, wipes, etc.) Feminine Care: Always products (pads, liners, etc.) Home Care - Cleaning: Flash Professional, Ariel products Home Care - Air Care: Febreze products Health Care: Pepto-Bismol, Clearblue products Other: Items that don't fit above categories Demo Products BITAM SOFT BITAM SOFT BABYLINO SENS.ECO. SpicyDeluxe CERAMIC CAP WELDRO PICKLING KINGS SAFETY SHOES ΠΕΤΡΕΛΑΙΟ ΘΕΡΜΑΝΣΗΣ double ,mint Instructions: P&G Product Identification: Use your knowledge of P&G brands and products to identify which receipt items are likely P&G products. Consider brand names, product types, and common P&G product categories but it is critical to never make up products. Category Assignment: Assign appropriate category based on the identified product type and brand. Identification Reasoning: Provide clear reasoning for why you identified each item as a P&G product or not. Field Handling: If any field is not clearly present in the text, omit it from the final JSON. Price Validation: Ensure all prices are in consistent format (e.g., "12.99" not "$12.99"). Decision Authority: Use your knowledge of P&G's brand portfolio to make identification decisions. Focus on recognizing P&G brands and typical product naming patterns. Enter the prompt text that will be used for image analysis.