Skip to main content

Prerequisites

Before starting any deployment method, ensure you have:

Required Tools​

  • kubectl configured and connected to your cluster
  • helm installed (version 3.x)
  • openssl for generating secure secrets

Infrastructure Requirements​

  • Installed and deployed AI/Run CodeMie instance
  • PostgreSQL database instance
  • Sufficient cluster resources for Langfuse components
  • Kubernetes cluster with appropriate permissions
  • Object storage bucket for Langfuse event, batch-export, and media uploads — one of:
    • AWS S3 (recommended for EKS): a dedicated S3 bucket with an IRSA role granting s3:PutObject, s3:GetObject, s3:DeleteObject, s3:ListBucket. Provisioned automatically via codemie-terraform-aws-platform with TF_VAR_enable_langfuse_s3=true.
    • Azure Blob Storage: a Storage Account with a container; static credentials required (IRSA not supported by Langfuse for Azure).
    • Google Cloud Storage: a GCS bucket. On GKE, use Workload Identity (recommended) — annotate the Langfuse service account with iam.gke.io/gcp-service-account and grant the GCP SA storage.objectAdmin on the bucket; no static credentials needed. Also can provide a service-account JSON key.
Minimum CodeMie Version

Minimal supported version of AI/Run CodeMie for evaluation is 1.3.0. Make sure you've updated your CodeMie installation before proceeding.

Access to the following repositories:​

Review Langfuse system requirements before proceeding.