NexusHR-AI

production-oriented HR assistant that uses Retrieval-Augmented Generation (RAG) to answer policy questions from uploaded documents (PDFs).

Backend (FastAPI)

• ✅ JWT Authentication with role-based access control
• ✅ RAG System with ChromaDB vector store
• ✅ Document Processing (PDF upload and parsing)
• ✅ Hugging Face Integration (embeddings + LLM)
• ✅ LangChain – RAG orchestration and retrieval pipeline

Frontend (Next.js 16)

• ✅ Modern UI with Tailwind CSS
• ✅ Real-time Chat Interface
• ✅ Document Management

Why this project

• Provide employees and HR teams with an intelligent assistant that answers policy questions using company documents (PDFs), and supports secure role-based access.

Features

• Secure authentication and role-based access (admin, hr_manager, employee)
• Document ingestion (PDF) with intelligent chunking and metadata
• Vector search with ChromaDB + sentence-transformers embeddings
• RAG-powered Q&A with context-aware retrieval and intent classification
• Web UI for chat and document management (Next.js + TypeScript)

Quick start (local)

Prerequisites

• Python 3.9+
• Node.js 18+ and npm
• Git

Backend

cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
cp .env.example .env
# Edit .env to add HUGGINGFACE_API_TOKEN and SECRET_KEY
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

Frontend

cd frontend
npm install
npm run dev

🔐 Demo Accounts

• Admin: hr_admin / admin123
• Manager: hr_manager / manager123
• Employee: employee / employee123

📚 API Documentation

• Swagger: http://localhost:8000/api/docs
• ReDoc: http://localhost:8000/api/redoc

🛠️ Tech Stack

• Backend: FastAPI, LangChain, ChromaDB, Hugging Face
• Frontend: Next.js 16, TypeScript, Tailwind CSS
• ML: sentence-transformers, google/flan-t5-large

---