Question

A retail expense-management startup wants to automatically extract structured fields (merchant name, date, line-item totals, tax) from photos of paper receipts that customers upload through a mobile app. The team is small, has no ML expertise, and wants the fastest path to a production-quality solution with minimal custom model training. Which Google Cloud service should they use?

Accepted Answer

Document AI Expense parser, a specialized pretrained processor that returns structured receipt fields directly

Answer

Cloud Vision API text detection (OCR) and then write custom regex parsing logic to identify each field

Answer

Train an AutoML Vision image classification model to categorize receipts by merchant type

Answer

Use BigQuery ML with a remote Gemini model to interpret raw receipt images stored as base64 strings