All Docs

Introduction to NFYio

Learn what NFYio is — a self-hosted cloud infrastructure platform with S3-compatible storage, AI/RAG pipelines, and multi-tenant architecture.

NFYio is a self-hosted cloud infrastructure platform that combines S3-compatible object storage with production-ready AI/RAG pipelines, enterprise-grade authentication, and multi-tenant architecture. Deploy it on your own infrastructure and retain full control over your data.

What is NFYio?

NFYio provides a complete cloud stack you can run anywhere:

  • Self-hosted cloud infrastructure — Deploy on-premises, in your data center, or any cloud provider
  • S3-compatible object storage — Drop-in replacement for AWS S3, backed by SeaweedFS
  • AI/RAG pipeline — Automatic embeddings, semantic search, and agentic workflows out of the box
  • Multi-tenant architecture — Workspace → Project → Team hierarchy with role-based access control

Key Features

Object Storage (S3 API)

Store and retrieve objects using the industry-standard S3 API. NFYio is compatible with AWS SDKs, aws CLI, and any S3-compatible tooling. Buckets support versioning, lifecycle policies, and fine-grained access control.

AI Agents (RAG, LLM, Workflow)

  • RAG (Retrieval-Augmented Generation) — Ingest documents, auto-generate embeddings, and query with semantic search
  • LLM integration — GPT-4o and other models via OpenAI-compatible APIs
  • Workflow runtime — Multi-step agentic pipelines with LangChain and policy gateway

VPC Networking

Isolate workloads with virtual private clouds, subnets, and security groups. NFYio supports network segmentation for multi-tenant deployments.

Embeddings (pgvector)

Documents are automatically embedded using OpenAI or Voyage AI models. Embeddings are stored in PostgreSQL with pgvector for fast similarity search at scale.

Security (Zero-Trust, MFA, Audit Logs)

  • Zero-trust authentication — Keycloak-based auth with JWT, MFA, CSRF protection
  • Session encryption — AES-256-GCM for session data
  • Audit logs — Track access and changes for compliance

Architecture

NFYio is built on a modular architecture:

LayerTechnology
StorageSeaweedFS — distributed object storage
DatabasePostgreSQL + pgvector for metadata and embeddings
AIOpenAI, Voyage AI — embeddings and LLM inference
AuthKeycloak — identity and access management
CacheRedis — sessions, queues, and caching
┌──────────────────────────────────────────────────┐
│                  NFYio Platform                  │
│                                                  │
│  ┌──────────┐  ┌──────────┐  ┌──────────────┐   │
│  │ Gateway  │  │ Storage  │  │    Agent     │   │
│  │ (auth,   │  │  Proxy   │  │   Service    │   │
│  │  teams,  │  │  (S3)    │  │  (RAG, LLM)  │   │
│  │ billing) │  └─────┬────┘  └──────┬───────┘   │
│  └────┬─────┘        │              │            │
│       │        ┌─────▼──────────────▼─────┐      │
│       └────────►       PostgreSQL          │      │
│                │    + pgvector             │      │
│                └─────────────┬─────────────┘      │
│                              │                   │
│  ┌────────┐  ┌───────────┐   │  ┌────────────┐  │
│  │ Redis  │  │ SeaweedFS │   │  │  Keycloak  │  │
│  │(queues)│  │  (blobs)  │   │  │   (auth)   │  │
│  └────────┘  └───────────┘   │  └────────────┘  │
└──────────────────────────────────────────────────┘

Use Cases

  • Document management with AI search — Ingest PDFs, docs, and images; search semantically across your corpus
  • RAG chatbots — Build private chatbots that answer from your own knowledge base
  • Private cloud — Replace or augment public cloud with self-hosted S3 and AI
  • Multi-tenant SaaS — Offer storage and AI services to multiple teams or customers with isolation

Next Steps

Edit this page on GitHub