kubeagi / arcadia

A diverse, simple, and secure one-stop LLMOps platform

Arcadia: A diverse, simple, and secure one-stop LLMOps platform

What is Arcadia?

Arcadia comes from Greek mythology(a tranquil and idyllic region, representing harmony, serenity, and natural beauty). We aim to help everyone find a more perfect integration between humans and AI.

To achieve this goal, we provide this one-stop LLMOps solution. Furthermore, we can easily host Arcadia at any Kubernetes cluster as production ready by integrating kubebb(Kubernetes building blocks).

Features

• Multi-tenant isolation (data, model services), built-in OIDC, RBAC, and auditing, supporting different companies and departments to develop through a unified platform
• Kubernetes native AGI agent orchestration
• Built on langchaingo(golang), has better performance and maintainability
• Support distributed inference using Ray
• Support quality and performance evaluation of AGI agent under different configurations
• A development and operational platform for AI agents, along with an AI agent portal for end-users
• Developed based on micro frontends and low-code approach, allowing for quick scalability and integration

Architecture

Our design and development in Arcadia design follows operator pattern which extends Kubernetes APIs.

For details, check Architecture Overview

Quick Start

Documentation

Visit our online documents

Read user guide

LLMs

List of supported(tested) LLMs

• baichuan2-7b
• chatglm2-6b
• qwen-7b-chat / qwen-14b-chat / qwen-72b-chat
• llama2-7b
• Mistral-7B-Instruct-v0.1
• bge-large-zh embedding
• m3e embedding
• ZhiPuAI(智谱 AI)
  • example
  • embedding
• DashScope(灵积模型服务)
  • example
  • text-embedding-v1(通用文本向量 同步接口)

Embeddings

Fully compatible with langchain embeddings

VectorStores

Fully compatible with langchain vectorstores

• ✅ PG Vector, KubeAGI adds the PG vector support to langchaingo project.
• ✅ ChromaDB

Pure Go Toolchains

Thanks to langchaingo,we can have comprehensive AI capability in Golang!But in order to meet our own unique needs, we have further developed a number of other toolchains:

• Optimized DocumentLoaders: optimized csv,etc...
• Extended LLMs: zhipuai,dashscope,etc...
• Tools: bingsearch,weather,etc...
• AppRuntime: powerful node(LLM,Chain,KonwledgeBase,vectorstore,Agent,etc...) orchestration runtime for arcadia

We have provided some examples on how to use them. See more details at here

CLI

We provide a Command Line Tool arctl to interact with arcadia. See here for more details.

• ✅ datasource management
• ✅ RAG evaluation

Contribute to Arcadia

If you want to contribute to Arcadia, refer to contribute guide.

Support

If you need support, start with the troubleshooting guide, or create GitHub issues