Amazon Bedrock 调用日志可观测性数据湖方案
原标题：Amazon Bedrock tokenomics data lake
AI 摘要
AWS 博客介绍了一套面向 Amazon Bedrock 的集中式可观测性数据湖方案。该方案通过 CloudWatch 日志订阅过滤器将各账号的模型调用日志汇聚到中心账号，经 Lambda 清洗、Firehose 转 Parquet 写入 S3，最终用 Athena 统一查询。它让平台和 ML 工程师能查看每次调用的提示词、响应、token 用量、模型和 IAM 身份，并可与成本报告关联，从而定位异常、识别浪费的提示词并按团队归因用量。
正文节选
Amazon Bedrock tokenomics data lake If you need an observability data lake for your Amazon Bedrock workloads, you’re not alone. When you’re running thousands of model invocations a day across multiple accounts, after a request is sent you might be wondering, what happened? You can see the aggregate bill, but you can’t see what happened inside each call. You might start to ask questions: what prompt was sent, what the model returned, how many tokens it used, which model handled it, and whether th