Frequently Asked Questions

Your AWS credentials (access keys, secret access keys, and session tokens) are stored exclusively in your browser's IndexedDB database. Before any credential touches the disk, it is encrypted using AES-256-GCM via the browser's native Web Crypto API.

Inspecting IndexedDB in your browser developer tools reveals only an opaque ciphertext blob. Zero plaintext keys or configuration metadata ever exist unencrypted on disk or on any server.

Amazon S3 Vectors is AWS's cloud storage tier designed for hosting large-scale vector embeddings, indexes, and vector buckets on high-durability Amazon S3 infrastructure.

Vector Studio provides a unified developer console allowing you to create vector buckets, create vector indexes with custom dimensions and distance metrics, inspect stored vector records, and run similarity queries in real time.

In vector databases, every index has a fixed vector dimension and space. Vectors generated by different embedding models occupy entirely different vector spaces—comparing an embedding from one model against an index built with another results in complete nonsense and dimension mismatch errors.

Vector Studio automatically locks the embedding provider to the single model compatible with that index:

• 256, 512, or 1024 Dimensions: Amazon Bedrock Titan Text Embeddings v2 (amazon.titan-embed-text-v2:0).
• Other Dimensions (e.g. 768D, 1536D): High-performance deterministic vector model tailored for the index's exact dimensions.