A manufacturing organization has a large collection of images labeled as intact or defective parts. They want to use this data to build a simple solution to detect faulty parts on their production lin

Pre-trained APIs use Google's generic models and cannot be trained on an organization's custom labeled dataset, so they would not recognize industry-specific defect patterns.

Document AI is designed for extracting information from structured and unstructured documents such as forms and invoices, not for classifying images of manufactured parts.

AutoML on Google Cloud enables users with no ML expertise to train high-quality custom models by uploading labeled training data, such as images tagged as intact or defective. It automates model architecture search and training, producing a custom model tailored to the organization's specific parts and defect types.

Discovery AI for Retail is a specialized service for product discovery and recommendations in e-commerce scenarios, not for manufacturing quality control image classification.