Skip to content

thomashousset/RegulAIte

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RegulAIte

Multi-agent AI system for Governance, Risk & Compliance (GRC)

Grounded in your own documents — not generic AI knowledge.

License: MIT Python React FastAPI Docker Qdrant

Lire en français


What is RegulAIte?

RegulAIte is an open-source GRC assistant built by students from OteriaCyberSchool. It lets organizations upload their internal documents (policies, standards, contracts) and query them through a conversational AI — powered by a multi-agent architecture and a RAG (Retrieval-Augmented Generation) pipeline.

Important

Instead of relying on generic AI knowledge, RegulAIte answers compliance questions based only on your documents, making responses traceable and auditable.


Features

Feature Description
🤖 Multi-agent orchestration An Orchestrator delegates tasks to specialized agents (compliance analysis, gap analysis, risk assessment, governance)
🔍 RAG pipeline Documents are chunked, embedded, and stored in a vector database (Qdrant) for precise retrieval
📄 Document management Upload and parse PDF, Word, and other formats via the Unstructured API
Async task queue Long-running agent tasks handled by Celery + Redis, keeping the UI responsive
🔐 Authentication User and organization management built-in
🌍 Bilingual support French and English query handling

Architecture

flowchart TD
    A7["Chat (Frontend)"]
    A9["Documents (Frontend)"]
    A8["Auth"]
    A6["API Routing (FastAPI)"]
    A5["Task Queue (Celery/Redis)"]
    A1["Orchestrator Agent"]
    A0["Specialized Agents"]
    A4["Tool Registry"]
    A2["RAG System (Qdrant)"]
    A3["Document Chunks"]

    A7 -- "Sends requests" --> A6
    A9 -- "Uploads documents" --> A6
    A8 -- "Protects" --> A6
    A6 -- "Routes to" --> A1
    A6 -- "Queues tasks" --> A5
    A5 -- "Runs" --> A0
    A5 -- "Processes" --> A3
    A1 -- "Delegates to" --> A0
    A0 -- "Uses" --> A4
    A0 -- "Queries" --> A2
    A4 -- "Queries" --> A2
    A2 -- "Retrieves" --> A3
Loading

Tech Stack

Layer Technology
🖥️ Frontend React
⚙️ Backend Python / FastAPI
🗄️ Vector DB Qdrant
🗃️ Relational DB MariaDB
📬 Task queue Celery + Redis
📑 Document parsing Unstructured
🐳 Containerization Docker Compose

Getting Started

Prerequisites

  • Docker and Docker Compose
  • An LLM API key (OpenAI, Anthropic, or compatible)

Run

git clone git@github.com:HXLLO/RegulAIte.git
cd RegulAIte

# Copy and configure environment variables
cp backend/.env.example backend/.env
# Edit backend/.env with your API keys and credentials

docker compose up --build

Once running, access the services:

Service URL
🌐 Frontend http://localhost:3000
🔌 Backend API http://localhost:8000
📖 API docs http://localhost:8000/docs
📊 Celery monitor http://localhost:5555
🔷 Qdrant dashboard http://localhost:6333/dashboard

Project Structure

RegulAIte/
├── backend/
│   ├── agent_framework/     # Agents, orchestrator, tools, RAG
│   ├── api/                 # FastAPI routes
│   ├── queuing_sys/         # Celery workers
│   └── config/              # Service configurations
├── front-end/               # React application
├── database-backups/        # Qdrant snapshots
├── docs/                    # Technical documentation
└── docker-compose.yml

Documentation

Detailed technical documentation is available in /docs:


Team

Built by students at OteriaCyberSchool — French cybersecurity engineering school.


License

MIT

About

No description, website, or topics provided.

Resources

Stars

4 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors