search

Self-Hosted Requirements

hashtag
Self-Hosted System Requirements

RootCause.ai is deployed via a single Helm chart into your Kubernetes environment. All components run within your infrastructure. If Kubernetes is not available, a bootstrap script is provided.

What’s Installed:

  • perceptura-platform: UI and backend-for-frontend (Next.js + API)

  • data-service: Core backend for data processing, orchestration, and LLM integration (FastAPI)

  • ml-jobs: Headless compute engine for batch and algorithmic workloads

Dependencies (bring your own or deploy via our charts):

  • MongoDB (8.0+)

  • PostgreSQL (16.0+)

  • Redis (7.2.4)

  • RabbitMQ (3.12.2+)

  • Temporal (1.30.0+)

  • MinIO or other S3-compatible storage

  • FusionAuth (optional)

  • LiteLLM for LLM integration

Infrastructure Requirements:

  • Kubernetes v1.30.0+, amd64 nodes

  • 3–4 nodes, each with 8 vCPU, 64GB RAM

  • SSD-backed persistent volumes

  • Optional: GPU (min 48GB VRAM) if running local LLMs

LLM Integration:

  • RootCause.ai prefers to integrate with your existing LLM infrastructure whenever possible (eg OpenAI, Azure OpenAI, Claude, private model).

  • If one is not available, RootCause.ai can optionally deploy with a local model on GPU nodes, exposed via an OpenAI-compatible API.

Customer Responsibilities:

  • Provision Kubernetes cluster with SSD-backed storage

  • Provide API keys/config for external services (databases, LLMs, etc.)

  • Set up TLS certificates and DNS for user access

  • Ensure outbound internet for telemetry, registry, and external LLMs

RootCause.ai Provides:

  • Complete Helm chart

  • Infrastructure guidance and architecture docs

  • Deployment support via Slack, email, and scheduled check-ins

PreviousUpgradingNextSSO Configuration

Last updated