Tokenomics at scale: How Jamf built real-time spend enforcement for Amazon Bedrock
Machine Learning Blog
This article describes how Jamf implemented real-time spend enforcement for Amazon Bedrock to control generative AI costs while maintaining developer productivity.
- Tracks daily per-user Amazon Bedrock spending using Athena cost views querying S3 logs
- Applies tiered model restrictions (denying premium models at 80-100% of daily budget) via IAM Customer Managed Policies
- Enforces restrictions within minutes using Lambda triggered every 15 minutes by EventBridge
- Maintains low-cost model access so engineers can continue working at budget limits
- Supports time-boxed exceptions via Slack slash commands with automatic TTL cleanup
- Costs under $10/month for hundreds of engineers; requires columnar log format to optimize Athena queries
The solution enables organizations to confidently expand AI access by providing observable, enforceable spend governance without blocking productivity.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Jul 28
2026
2026
Implementing per-user token guardrails for Amazon Bedrock in government agencies
Sep 1
2026
2026
How t54 built a trust layer with Amazon Bedrock AgentCore payments
Aug 18
2026
2026
Amazon Bedrock AgentCore payments is now generally available: Enabling agents to transact safely and autonomously at scale
Sep 1
2026
2026
Human-in-the-loop claims processing with Amazon Bedrock AgentCore
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.