New Region support for Guardrails : Amazon Bedrock Guardrails is now supported in Asia Pacific (Jakarta).

Amazon Bedrock Release Notes

Amazon Bedrock is a fully managed service that offers high-performing foundation models from leading AI companies and Amazon, accessible through a unified API. It allows you to build generative AI applications with security, privacy, and responsible AI practices.

What can I do with Amazon Bedrock?

• Experiment with prompts and configurations: Generate responses by submitting prompts and using different configurations and foundation models.
• Augment response generation: Create knowledge bases by uploading data sources to enhance response generation.
• Build agents: Develop agents that use foundation models to reason through and carry out tasks for customers.
• Customize models: Fine-tune or continue pre-training models with your data to improve performance on specific tasks or domains.
• Purchase Provisioned Throughput: Run inference on models more efficiently and at discounted rates.
• Evaluate models: Determine the best model for your use case by evaluating outputs with built-in or custom prompt datasets.
• Use guardrails: Implement safeguards for your generative AI applications to prevent inappropriate or unwanted content.
• Optimize latency: Get faster response times with Latency-optimized inference for foundation models.

Note: The Latency Optimized Inference feature is in preview release and subject to change.

What to do

• Familiarize yourself with the terms and concepts used in Amazon Bedrock.
• Understand how AWS charges for using Amazon Bedrock.
• Try the Getting started with Amazon Bedrock tutorials.
• Read the documentation for the features you want to include in your application.

Source: AWS release notes