Q11 — AWS SAA-C03 Ch.15

Question 11 of 100 | ← Chapter 15

Q1111. A company launches a new web application that uses an Amazon Aurora PostgreSQL database. The company wants to add new features to the application that rely on AI. The company requires vector storage capability to use AI tools.Which solution will meet this requirement MOST cost-effectively?

• A. Use Amazon OpenSearch Service to create an OpenSearch service. Configure the application to write vector embeddings to a vector index.
• B. Create an Amazon DocumentDB cluster. Configure the application to write vector embeddings to a vector index.
• C. Create an Amazon Neptune ML cluster. Configure the application to write vector embeddings to a vector graph.
• D. Install the pgvector extension on the Aurora PostgreSQL database. Configure the application to write vector embeddings to a vector table. ✓

Correct Answer: D. Install the pgvector extension on the Aurora PostgreSQL database. Configure the application to write vector embeddings to a vector table.

Explanation

To determine the most cost-effective solution for adding vector storage capability to an Amazon Aurora PostgreSQL database for an AI-driven web application, let's evaluate each option:A. Use Amazon OpenSearch Service to create an OpenSearch service:Amazon OpenSearch Service provides a vector engine for scalable, secure, and high-performance vector databases.It supports advanced k-NN and ANN algorithms, as well as HNSW and IVF indexes, for efficient vector search.However, it is a separate service from Aurora PostgreSQL, which may involve additional costs and integration efforts.B. Create an Amazon DocumentDB cluster:Amazon DocumentDB is a MongoDB-compatible document database that can support vector storage through extensions or custom implementations.It offers scalability and performance suitable for large-scale datasets. However, it is also a separate service, requiring additional setup and maintenance costs.C. Create an Amazon Neptune ML cluster:Amazon Neptune ML is designed for graph-based machine learning and is not primarily intended for vector storage.It may not be the most suitable choice for this specific requirement, as it focuses on graph data rather than vector embeddings.D. Install the pgvector extension on the Aurora PostgreSQL database:The pgvector extension allows Aurora PostgreSQL to store, query, and index vector data directly within the database.This eliminates the need for a separate vector database, reducing costs and simplifying integration. Aurora PostgreSQL, being a managed service, already offers high performance and scalability, making it a suitable platform for vector storage.Given these considerations, the most cost-effective solution would be:D. Install the pgvector extension on the Aurora PostgreSQL database and configure the application to write vector embeddings to a vector table.This option leverages the existing Aurora PostgreSQL infrastructure, minimizing additional costs and integration complexity. It also benefits from Aurora's managed services, ensuring high performance and scalability.