- Adds Intelligent Ingestion to OpenSearch Serverless, automatically parsing and indexing logs and documents from Amazon S3.
- Supports structured (JSON, NDJSON) and unstructured (PDF, Word, plain text, Markdown) formats, generates vector embeddings and vision‑language descriptions, and requires no ingestion infrastructure.
- Configurable via console or API, scales with workload, pay‑only‑for‑processing, and works across accounts and regions.