AWS Machine Learning Describes an Amazon Bedrock Observability Data Lake

AWS Machine Learning described a centralized observability data lake for Amazon Bedrock invocations in a blog post. The design pulls invocation logs from linked AWS accounts into a central account, processes them with Amazon Data Firehose, and exposes a single queryable table in Amazon Athena. The setup lets teams inspect prompts, responses, token counts, model choice, and per-invocation cost. AWS also provided CloudFormation templates, Lambda code, and SQL queries in a sample GitHub repository.