- AWS Elemental Inference now supports customization using Amazon Bedrock multi-modal models for open-ended custom questions or contextual metadata decisions.
- Customizations run on the existing pipeline and return results through the GetMetadata API without requiring a separate infrastructure.