How BMW Group detects cost anomalies across 14,000 cloud accounts
BMW Group's CLEA system runs daily anomaly detection across 14,000 cloud accounts for ~$50/month in compute.
“But a dashboard shows what already happened, and only when someone opens it.”
BMW Group and Reply built CLEA, a serverless FinOps pipeline on AWS that uses Meta's Prophet library to forecast expected spend and alert account owners to cost anomalies across 14,000 cloud accounts. It is a practical ML-for-cost-management engineering case study rather than a major AI industry signal, of limited relevance to conference-level AI intelligence.