https://github.com/qwerryfiwy/neurax
NeuraX is an AI assistant powered by AzureOpenAI.
https://github.com/qwerryfiwy/neurax
ai azureopenai dotenv fastapi framer-motion nextjs python tailwindcss
Last synced: 2 months ago
JSON representation
NeuraX is an AI assistant powered by AzureOpenAI.
- Host: GitHub
- URL: https://github.com/qwerryfiwy/neurax
- Owner: qwerryfiwy
- License: mit
- Created: 2025-06-21T05:39:47.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-12-12T13:21:06.000Z (6 months ago)
- Last Synced: 2025-12-14T04:12:03.736Z (6 months ago)
- Topics: ai, azureopenai, dotenv, fastapi, framer-motion, nextjs, python, tailwindcss
- Language: TypeScript
- Homepage: https://neura-x-eight.vercel.app
- Size: 9.44 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# NeuraX
> An AI-powered Deep Learning assistant + smart document brain, built for coders and students.
---
## Overview
**NeuraX** is a smart, domain-specific AI assistant that helps users:
- Understand Deep Learning concepts
- Upload PDFs and get instant summaries
- Ask contextual questions grounded in your documents
It's built with the power of **Azure OpenAI (GPT-4o)** and a blazing fast **FastAPI backend**, combined with a slick **Next.js frontend** styled in TailwindCSS.
---
## Modules Included
### 1. NeuraX Assistant
Ask anything related to Deep Learning:
- L1 vs L2 regularization
- What is vanishing gradient?
- CNN vs RNN vs Transformer
- Backpropagation, optimizers, activation functions...
NeuraX only answers DL-related queries using prompt control. No off-topic distractions.
---
### 2. NeuraX Docs
Upload your notes or research papers (`.pdf`) and:
- Automatically extract and summarize the content
- Ask questions based only on your uploaded doc
- Uses embeddings + FAISS to fetch relevant text chunks
Works offline, remembers your doc, and responds fast.
---
## Tech Stack
| Layer | Tech Used |
|-------------|---------------------------------------------------------------------------|
| Frontend | [Next.js](https://nextjs.org/docs), [TailwindCSS](https://tailwindcss.com/docs), [Framer Motion](https://www.framer.com/motion/), [TypeScript](https://www.typescriptlang.org/docs/) |
| Backend | [FastAPI](https://fastapi.tiangolo.com/), [Python](https://docs.python.org/3/), [Azure OpenAI GPT-4o](https://learn.microsoft.com/en-us/azure/cognitive-services/openai/overview) |
| Embedding | [Azure OpenAI Embeddings](https://learn.microsoft.com/en-us/azure/cognitive-services/openai/how-to/use-embeddings) |
| Vector DB | [FAISS](https://github.com/facebookresearch/faiss) + [Pickle](https://docs.python.org/3/library/pickle.html) for local persistence |
| Security | [.env-based Key Management](https://12factor.net/config) ([dotenv for Python](https://pypi.org/project/python-dotenv/), [dotenv for Next.js](https://nextjs.org/docs/basic-features/environment-variables)) |
---
## Folder Structure
NeuraX/
├── frontend/ # Next.js frontend
│ ├── pages/assistant.tsx
│ ├── pages/docs.tsx
│ └── components/
├── backend/ # FastAPI backend
│ ├── main.py # Unified backend
│ ├── vector_db/ # FAISS index + chunk data
│ └── .env
## API Endpoints
### `POST /ask`
> Deep Learning assistant
```json
{"prompt": "Explain dropout in neural nets"}
```
### POST /upload-doc
Upload a PDF and get a summary
{
"summary": "This paper covers gradient descent and..."
}
## Setup Instructions
### Requirements
-Node.js, npm
-Python 3.9+
-Azure OpenAI access (GPT-4o + Embedding deployment)
## Frontend
---
## Setup Instructions
### Frontend
```bash
cd frontend
npm install
npm run dev
```
### Backend
```bash
cd backend
python -m venv venv
source venv/bin/activate # For Windows: venv\Scripts\activate
pip install -r requirements.txt
AZURE_API_KEY=your-azure-key
AZURE_OPENAI_ENDPOINT=https://your-resource-name.openai.azure.com/
uvicorn main:app --reload --port 8001
```