QuestionQ126

VerifiedOperational Efficiency and Optimization for GenAI Applications

A retail company operates a product-recommendation system that uses a foundation model (FM) in Amazon Bedrock. The system processes and searches millions of product embeddings, which are updated hourly. The company regularly adds new products.

During peak shopping events, the current solution has high latency and occasional failures when it performs similarity searches across the product catalog. The company must address these performance issues.

Which solution will meet this requirement?

Explanation

Amazon OpenSearch Service supports k-nearest-neighbor vector search for embeddings, including approximate k-NN methods that improve search performance at large scale. A multi-node architecture distributes vector indexing and search workloads, while k-NN index settings, memory safeguards, and request-queue capacity help sustain high query volume and avoid resource-related failures during peak demand.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!