Orbicorp - AI Agent Management Platform

An enterprise-grade AI agent management platform. Multi-tenant, multi-model, enterprise-ready.

Introduction

Orbicorp is an AI agent platform built for small and medium-sized businesses (SMBs) that want the productivity of custom AI assistants without the cost and complexity of building one from scratch. Instead of juggling separate chatbots, automation tools, and integrations, your team manages a fleet of purpose-built AI agents from a single, multi-tenant dashboard.

Each agent can be tailored to a specific role — customer support, sales follow-up, HR onboarding, internal knowledge lookup — and put to work across the channels your customers already use.

Key Features

• Custom AI agents, no code required — Create, configure, pause, and manage role-specific agents (support, sales, HR, operations) from a simple web dashboard.
• Bring your own model, avoid lock-in — Connect Anthropic (Claude), OpenAI (GPT), Google (Gemini), or 200+ models through a single OpenRouter key, and switch providers anytime to control cost and quality.
• Multi-channel customer engagement — Reach customers where they are with built-in Telegram and WhatsApp messaging, plus a real-time streaming web chat.
• Knowledge base (RAG) — Ground agents in your own documents and company data so answers stay accurate and on-brand instead of generic.
• Agent memory — Agents remember facts, preferences, and past interactions, delivering a personalized experience that improves over time.
• Multi-agent collaboration — Agents can delegate tasks to one another and a coordinator agent, automating multi-step workflows end to end.
• Task & workflow automation — Turn repetitive business processes into tracked tasks and automated workflows that run without manual follow-up.
• Skills marketplace — Extend agents with installable skills, tools, language packs, and ready-made templates to launch faster.
• Team & role management — Invite teammates with role-based access (Admin, Operator, Viewer) and keep each company's data fully isolated.
• Insights dashboard — Track usage, activity, and agent performance to understand where AI is saving your team time.
• Localized & accessible — Turkish/English interface, dark/light themes, and a mobile-friendly responsive design out of the box.

Why SMBs choose Orbicorp

• Fast to launch — Sensible defaults, seed data, and templates get you running in minutes, not months.
• Affordable & flexible — Pay only for the models you use and scale agents up or down as your needs change.
• Self-hostable — Runs on your own infrastructure with Docker (PostgreSQL + Redis), keeping your customer data under your control.

🏗️ Project Structure

orbicorp-full/
├── orbicorp-server/        # Backend (Node.js + Fastify + Prisma)
│   ├── src/
│   │   ├── modules/
│   │   │   ├── auth/           # JWT authentication
│   │   │   ├── agents/         # Agent CRUD
│   │   │   ├── sessions/       # Chat sessions
│   │   │   ├── users/          # User management
│   │   │   ├── companies/      # Company settings
│   │   │   └── agent-runtime/  # LLM orchestration
│   │   └── ...
│   ├── prisma/                 # Database schema
│   └── docker-compose.yml      # PostgreSQL + Redis
│
└── orbicorp-frontend/      # Frontend (Static HTML/CSS/JS)
    ├── index.html              # Main shell
    ├── orbicorp-*.html         # Page modules
    ├── orbicorp-api.js         # API client
    └── ...

🚀 Quick Start

1. Backend Setup

cd orbicorp-server

# Environment variables
cp .env.example .env
# Edit the .env file (JWT secrets, LLM API keys)

# Start PostgreSQL + Redis with Docker
docker-compose up -d

# Install dependencies
npm install

# Prepare the database
npm run db:generate
npm run db:migrate
npm run db:seed

# Start the server
npm run dev

2. Frontend Setup

cd orbicorp-frontend

# Start an HTTP server
npx serve . -p 8000

# or
python3 -m http.server 8000

3. Open in the Browser

• Frontend: http://localhost:8000
• Backend API: http://localhost:3001/api/v1
• Swagger Docs: http://localhost:3001/docs

Demo User

Email:    admin@novatech.com.tr
Password: admin123

🤖 LLM Configuration

Add your API key to the .env file:

# Option 1: OpenRouter (RECOMMENDED - 200+ models with a single key)
OPENROUTER_API_KEY="sk-or-v1-..."

# Option 2: Direct provider keys
ANTHROPIC_API_KEY="sk-ant-..."
OPENAI_API_KEY="sk-..."
GOOGLE_API_KEY="..."

Supported Models

Provider	Models
OpenRouter	Claude, GPT, Gemini, Llama, Mistral, DeepSeek, Qwen...
Anthropic	Claude Opus 4.6, Sonnet 4.6, Haiku 4.5
OpenAI	GPT-4o, GPT-4o Mini, GPT-4 Turbo
Google	Gemini 1.5 Pro, Gemini 1.5 Flash

📡 API Endpoints

Auth

POST /api/v1/auth/register   - Register
POST /api/v1/auth/login      - Login
POST /api/v1/auth/refresh    - Refresh token
POST /api/v1/auth/logout     - Logout
GET  /api/v1/auth/me         - Current user info

Agents

GET    /api/v1/agents        - List
POST   /api/v1/agents        - Create
GET    /api/v1/agents/:id    - Details
PUT    /api/v1/agents/:id    - Update
DELETE /api/v1/agents/:id    - Delete

Sessions & Chat

POST   /api/v1/sessions              - Start a session
GET    /api/v1/sessions/:id/messages - Get messages
POST   /api/v1/sessions/:id/messages - Send a message
GET    /api/v1/sessions/:id/chat/stream - Streaming (SSE)

LLM

GET  /api/v1/llm/status      - Provider status
GET  /api/v1/llm/providers   - Model list
POST /api/v1/llm/test        - Connection test
POST /api/v1/llm/benchmark   - Comparison

🎨 Frontend Features

• Multi-tenant: Per-company isolation
• Dark/Light Mode: Theme support
• i18n: Turkish/English
• Responsive: Mobile-friendly
• Real-time: Streaming chat via SSE

🛠️ Tech Stack

Backend

• Runtime: Node.js 20+
• Framework: Fastify
• ORM: Prisma
• Database: PostgreSQL
• Cache: Redis
• Auth: JWT + Refresh Token

Frontend

• UI: Vanilla HTML/CSS/JS
• Icons: Lucide
• Styling: Custom CSS + Dark mode

📋 Roadmap

• Knowledge Base (RAG) integration
• Workflow Engine
• WebSocket notifications
• File upload / S3 integration
• Slack/Teams integration

📄 License

Proprietary - All rights reserved