Part 2: Amazon Bedrock cost attribution with Amazon Athena and CUDOS
Learn how to visualize and analyze Amazon Bedrock cost attribution using Amazon Athena and CUDOS dashboards. This post shows how to set up CUR 2.0 with IAM principal data, query Bedrock spend by principal, project, and team, and build dashboards to track AI costs across your organization.
Part 2 focuses on utilizing Amazon Athena and CUDOS for detailed Amazon Bedrock cost attribution. This section starts with the setup of Cost and Usage Reports (CUR 2.0) through Data Exports, specifically designed to include IAM principal data. The prerequisites for this setup include an AWS account with billing console access, necessary IAM permissions for Cost and Usage Reports, S3, and Athena, an S3 bucket for CUR data storage, and basic familiarity with SQL and the AWS Management Console.
An optional tool, Claude Code or Kiro-CLI, can automate the setup process if desired. It’s crucial to ensure IAM principal data is included in the CUR 2.0 export, as this step is required for accurate cost attribution. The setup involves selecting options in the export configuration such as enabling caller identity (IAM principal) allocation data and choosing hourly granularity for maximum detail.
The installation of this data can increase file sizes, so storage planning and lifecycle policies are recommended for high-volume workloads. After enabling IAM principal data, it takes up to 24 hours for AWS to deliver the first CUR 2.0 report to the designated S3 bucket. Next, the connection between CUR 2.0 and Amazon Athena is established, allowing for SQL-based querying without needing additional infrastructure management.
An optional agent.md skill repo is provided for this process, which includes a sample test query example in the Athena Query Editor. This query is designed to retrieve IAM principal ARNs and Bedrock usage types, confirming a successful setup. The final part of Part 2 introduces query patterns in Amazon Athena for Bedrock cost tracking.
The first query pattern breaks down Bedrock spending by IAM principal and usage type, answering questions about who is using which models and the associated costs. The sample query shown returns granular details of costs by IAM principal and model usage, enabling detailed cost analysis.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
