An e-commerce company manages a digital catalog of consumer products submitted by third-party sellers. Each product submission includes a description stored as a text file in an Amazon S3 bucket. These descriptions may include ingredient information for consumable products like snacks, supplements, or beverages. The company wants to build a fully automated solution that extracts ingredient names from the uploaded product descriptions and uses those names to query an Amazon DynamoDB table, which returns precomputed health and safety scores for each ingredient. Non-food items and invalid submissions can be ignored without affecting application logic. The company has no in-house machine learning (ML) experts and is looking for the most cost-effective solution with minimal operational overhead. Which solution meets these requirements MOST cost-effectively?