---
format: "aidr-story-markdown/v1"
id: "c2e46806d0d494d9f32dd546141cc934fb7b913c5629155ebcae674747883248"
canonical_url: "https://aidr.today/c2e46806?lang=en"
title: "How BMW Group detects cost anomalies across 14,000 cloud accounts"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-09-21T16:36:10.000Z"
category: "Models"
topics: ["anthropic","claude","multi-agent","open-source"]
source_urls: ["https://aws.amazon.com/blogs/machine-learning/how-bmw-group-detects-cost-anomalies-across-14000-cloud-accounts/"]
summary: "BMW Group operates CLEA, a FinOps platform monitoring more than 14,000 cloud accounts. This post shows how BMW added automated daily cost anomaly detection, moving from reactive dashboards to proactive alerts using Prophet forecasting, AWS Step Functions, and a serverless pipeline that processes every account for about $50 per month."
---

# How BMW Group detects cost anomalies across 14,000 cloud accounts

> [Open the canonical story](<https://aidr.today/c2e46806?lang=en>)

**Published:** 2026-09-21T16:36:10.000Z
**Category:** Models
**Topics:** anthropic, claude, multi\-agent, open\-source

## Summary

BMW Group operates CLEA, a FinOps platform monitoring more than 14,000 cloud accounts\. This post shows how BMW added automated daily cost anomaly detection, moving from reactive dashboards to proactive alerts using Prophet forecasting, AWS Step Functions, and a serverless pipeline that processes every account for about $50 per month\.

## Sources

- [Story source](<https://aws.amazon.com/blogs/machine-learning/how-bmw-group-detects-cost-anomalies-across-14000-cloud-accounts/>)