Skip to content

ECO-AIM-AI-011

Name: Storing all embeddings indefinitely

Category: AIM

Family: AI

Primary layer: ai

System layers: ai

Description

Embedding stores without retention grow unbounded and expensive.

Impact

  • confidence: 0.7
  • notes: Often overlooked.
  • type: storage

Detection

  • languages:
  • infra
  • method: config

Remediation

  • guidance: Apply retention and tiering; delete stale embeddings.
  • tradeoffs: Must re-embed when needed.

Pattern examples

No pattern examples provided.

Remediation examples

No remediation examples provided.