Eidolon -- AI Agent Server for the Enterprise

AI Agent Server for the Enterprise

An Enterprise-Grade, Open Source framework, Eidolon enables developers to rapidly build and deploy the most powerful genAI applications.

Get Started Star on GitHub ⭐️

What is Eidolon? The industry's first AI Agent Server for the Enterprise

Agent Server

A secure, enterprise ready, deployment server for Agentic applications.

Eidolon SDK

A pluggable AI Agent framework (SDK) for developers to build Agentic applications quickly and easily.

Bring agents into production in 3 easy steps

1. Build

2. Deploy

3. Consume

1. Build

Defining agents is fast and easy.

Choose from a variety of pre-built agents or define your own

Eidolon provides a variety of pre-built agents to get you started. You can also define your own agents using vanilla code or any other agent framework you like.

Define agents with declarative YAML

Eidolon's simple declarative YAML syntax makes it easy to define agents. Agents can be as simple as a single line or as complex as needed.

Build agents that work together

Add RAG storage to any application using simple services. Eidolon provides the flexibility to customize RAG to fit your needs.

apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: multi-model-chatbot

spec:
  system_prompt:
    You are a helpful assistant. Output all multi-media content as embedded content including image, video, and audio. Only generate images or speech if directly asked by the user.
  apus:
    - { apu: GPT4-turbo, title: GPT4 Turbo, default: true }
    - { apu: MistralLarge, title: Mistral Large }
    - { apu: Llama3-8b, title: Llama3 8b }
    - { apu: ClaudeOpus, title: Claude Opus }
    - { apu: ClaudeSonnet, title: Claude Sonnet }

---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: agent

spec:
  description: An agent that searches the eidolon code and documentation
  agent_refs: ["doc-search"]
  system_prompt: You are a software expert responsible for answering questions about a code base.
---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: doc-search

spec:
  implementation: RetrieverAgent
  description: "Search the Eidolon github repository for code, examples, and documentation"
  document_manager:
    loader:
      implementation: GitHubLoader
      owner: "eidolon-ai"
      repo: "eidolon"
      pattern: [ "**/*.md", "**/*.py" ]
      exclude: "**/test/**/*"

---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: manager

spec:
  system_prompt:
    You are a tech manager servicing user requests. You manage a team of engineers and qa workers. Work with your team as needed to answer user questions.
  agent_refs: ["engineer", "qa"]
---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: manager

spec:
  system_prompt:
    You are a tech manager servicing user requests. You manage a team of engineers and qa workers. Work with your team as needed to answer user questions.
  agent_refs: ["engineer", "qa"]
---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: engineer

spec:
  system_prompt:
    You are an engineer. Your manager (user) will come to with questions. Work with the qa team as needed to answer user questions.
  agent_refs: ["qa"]
---
apiVersion: server.eidolonai.com/v1alpha1
kind: Agent
metadata:
  name: qa

spec:
  system_prompt: You are a qa engineer. Quality is your top concern.

2. Deploy

What good is an agent that can't be deployed?

Deploy directly to Kubernetes

Agents are infrastructure and should be deployed as such. Deploy directly using your organizations Kubernetes management tools / pipeline.

Scale agents as needed

Agents are stateless and scale horizontally to meet your needs.

Build safely with enforceable policies

With kubernetes deployment, you can enforce policies to ensure your agents are deployed with access only to the resources they need.

% helm repo add eidolon https://eidolonai.com/charts
...
% helm install eidolon eidolon-operator/eidolon-operator-chart
...
% kubectl apply -f agent.eidolon.yaml
agent.server.eidolonai.com/agent created
% kubectl get agents
NAME      AGE
chatbot   3s

3. Consume

GenAI applications need to consume agents in a diverse set of ways.

Build a GUI with React components

Eidolon provides a set of React components to help you build a GUI to interact with your chatbot agents. The components are designed to be easy to use and extensible.

Run actions directly through HTTP REST API

At the end of the day, agents are just http services. You can run actions directly through the REST API or use the Eidolon webui to interact with your agents.

Hack away with the CLI

Eidolon provides a CLI to help you interact with your agents in a more ad-hoc manner. The CLI is a powerful tool to help you experiment with your agents quickly.

from eidolon_ai_client.client import Agent

agent = Agent.get("HelloWorld")
process = await agent.create_process()
body = "Why should I use Eidolon?"
response = await process.action("converse", body)

assert "fast iteration times" in response.data
assert "extensible agent mesh" in response.data
assert "supportive community" in response.data

import {EidolonClient} from "@eidolon-ai/client";

const client = new EidolonClient()
const {process: p1} = await client.createProcess("HelloWorld")
const process = client.process(p1.process_id)
const agent = process.agent("HelloWorld")
const body = "Why should I use Eidolon?"
const response = await agent.action("converse", body)

expect(response).toContain("fast iteration times")
expect(response).toContain("extensible agent mesh")
expect(response).toContain("supportive community")