LLMOps - End-to-End LLM Operations Platform

LLMOps is a full-stack platform for building and operating AI agent applications. The repository contains:

• A Flask backend with LangChain and LangGraph orchestration
• A Vue 3 frontend for agent, workflow, dataset, tool, and conversation management
• Celery workers for background jobs
• PostgreSQL, Redis, Weaviate, and Nginx in the Docker stack

The current codebase focuses on:

• Multi-provider LLM integration
• Workflow authoring and execution
• Conversation management and search
• Public app and workflow publishing
• Knowledge base, document, and dataset management
• Built-in tools and notification pipelines

.
├── api/        # Flask backend, services, handlers, tasks, tests
├── ui/         # Vue 3 frontend, components, views, tests
├── docker/     # Docker Compose stack and deployment config
├── docs/       # High-level documentation index and deployment guides
└── README.md   # Project overview

• REST APIs built with Flask
• Service-oriented backend structure
• JWT, OAuth, and role-based account flows
• SSE / websocket driven real-time notifications
• Celery-based background tasks

• Vue 3 + Vite + TypeScript
• Workflow editor and app management views
• Conversation history, search, and publishing pages
• Notification components and live UI updates

• Docker Compose deployment
• PostgreSQL for persistence
• Redis for cache and task queue
• Weaviate for vector search
• Nginx as reverse proxy

┌─────────────────────────────────────────────────────────────┐
│                         Frontend (Vue 3)                     │
│  Agent Builder  Workflow Editor  Dataset Manager  Tools UI   │
└─────────────────────────────────────────────────────────────┘
                            ↕ HTTP / SSE / WebSocket
┌─────────────────────────────────────────────────────────────┐
│                      Backend (Flask + Celery)                │
│  API Layer   Services   LangChain / LangGraph   Background   │
└─────────────────────────────────────────────────────────────┘
                            ↕
┌─────────────────────────────────────────────────────────────┐
│                    Infrastructure                            │
│  PostgreSQL   Redis   Weaviate   Nginx                       │
└─────────────────────────────────────────────────────────────┘

• Docker 20.10+
• Docker Compose 2.0+
• 8GB+ RAM recommended for the full stack

git clone https://github.com/Haohao-end/LMForge-End-to-End-LLMOps-Platform-for-Multi-Model-Agents.git
cd LMForge-End-to-End-LLMOps-Platform-for-Multi-Model-Agents

cp api/.env.example api/.env
# edit api/.env and fill in the required API keys

cd docker
docker compose up -d --build

For more deployment details, see DOCKER_QUICKSTART.md and docker/README.md.

cd api
pip install -r requirements.txt
flask run --port 5001

Run tests:

cd api
pytest

cd ui
npm install
npm run serve

Vite serves the frontend on port 5173 by default and proxies /api to http://localhost:5001.

Useful frontend commands:

cd ui
npm run type-check
npm run lint
npm run build
npm run test:unit -- --run

Copy api/.env.example to api/.env and set at least one LLM provider key plus the required database and Redis settings.

If you need to customize ports, container passwords, or other infrastructure settings, use the Docker configuration documented in docker/README.md.

• api/.env.example
• docker/README.md
• docker/SECURITY.md
• CLAUDE.md

• docs/deployment/DEPLOYMENT.md - deployment guide
• docs/deployment/QUICKSTART_GUIDE.md - quick start guide
• api/README.md - backend notes
• ui/README.md - frontend notes
• docker/README.md - Docker setup
• docker/SECURITY.md - security guidance

The repository already includes a large automated test suite.

• Backend: cd api && pytest
• Frontend: cd ui && npm run test:unit -- --run

License terms are not declared in a root LICENSE file at the moment. Add one if you want the licensing to be explicit.