Private Cloud AI Platform
Distributed ARM64 Kubernetes cluster powering AI inference, vector search, object storage, CI/CD automation, and self-hosted backend services.
Core Systems
A self-hosted distributed platform combining Kubernetes, AI inference, data systems, storage, and automation on ARM64 infrastructure.
AI Systems
LLM inference, embeddings, RAG pipelines, and agent workflows running on-cluster.
Data Layer
PostgreSQL, Redis, Weaviate vector search, Apache AGE graph engine, and search indexing.
Storage
MinIO S3-compatible object storage for datasets, model weights, and media assets.
System Architecture Diagram
Visual representation of compute, storage, AI, and backend layers across the cluster.
Loading diagram…
Explore by System
Each domain is independently documented with architecture, design decisions, and component details.
AI Systems
RAG · Agents · Inference · Embeddings
LLM inference pipeline, vector retrieval, semantic search, and agent orchestration on-cluster.
Backend Systems
APIs · Async · Services · Databases
FastAPI services, Redis caching, PostgreSQL with Apache AGE graph extension, job queues.
Infrastructure
Kubernetes · ARM64 · Homelab · CI/CD
k3s cluster across 8 Raspberry Pi 5 nodes with Gitea CI/CD, Prometheus, and Grafana.