A company has a large unstructured dataset containing many duplicate records across key attributes. Which AWS solution will identify duplicates with the least amount of custom coding?

The AWS Glue FindMatches transform is specifically designed for identifying duplicate or matching records in datasets, even when exact matches don't exist (fuzzy matching). It uses machine learning to learn patterns for matching records, requiring minimal custom coding. Amazon Mechanical Turk is a crowdsourcing service, which, while capable of identifying duplicates, involves significant human effort and management, making it less automated than the FindMatches transform. Amazon QuickSight ML Insights focuses on anomaly detection and forecasting within structured data for business intelligence, not general-purpose data deduplication. Amazon SageMaker Data Wrangler is a data preparation tool that can be used for preprocessing, but it doesn't offer a built-in, low-code solution for fuzzy duplicate detection like AWS Glue FindMatches. While you could potentially build a custom deduplication model in SageMaker, it would require significantly more custom coding than using FindMatches.