https://github.com/miashraf1818/genai-rag-chatbot

🤖 AI-powered RAG chatbot: Upload documents, ask questions, get intelligent answers. FastAPI + LLaMA 3.1 + Pinecone + Nextjs.
https://github.com/miashraf1818/genai-rag-chatbot

ai chatbot fastapi lang llama pinc python rag streamlit

Last synced: about 2 months ago
JSON representation

🤖 AI-powered RAG chatbot: Upload documents, ask questions, get intelligent answers. FastAPI + LLaMA 3.1 + Pinecone + Nextjs.

Host: GitHub
URL: https://github.com/miashraf1818/genai-rag-chatbot
Owner: miashraf1818
Created: 2025-11-13T19:14:58.000Z (7 months ago)
Default Branch: main
Last Pushed: 2026-01-13T10:02:58.000Z (5 months ago)
Last Synced: 2026-01-13T10:58:48.382Z (5 months ago)
Topics: ai, chatbot, fastapi, lang, llama, pinc, python, rag, streamlit
Language: TypeScript
Homepage:
Size: 337 KB
Stars: 1
Watchers: 1
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md

Awesome Lists containing this project

README

          # 🤖 GenAI RAG Chatbot

A production-ready, full-stack AI chatbot application featuring **Retrieval-Augmented Generation (RAG)**, real-time streaming, and secure user authentication. Built with modern "DevOps" practices using Docker and AWS.

## ✨ Key Features

### 🧠 Advanced AI Capabilities

This project is a RAG (Retrieval-Augmented Generation) chatbot that allows users to upload documents (PDF, DOCX, TXT) and chat with them using an LLM (Llama 3 via Groq). It features a modern, responsive UI built with Next.js and a robust FastAPI backend.

> **Latest Update:** Optimized for AWS EC2 t2.micro deployment (CPU-only inference).

-   **Smart Context**: Uses **Pinecone** vector database to retrieve relevant info for every query.

-   **Llama 3 Powered**: Integrated with **Groq** for lightning-fast AI responses.

-   **Streaming Responses**: Real-time text generation (like ChatGPT) using WebSockets.

### 🔒 Security & Auth

-   **Google OAuth 2.0**: Secure "Login with Google" integration.

-   **User Isolation**: Each user only sees *their* own documents and chat history.

-   **JWT Sessions**: Stateless, secure session management.

### 💻 Modern UI/UX

-   **React + TypeScript**: Fast, type-safe frontend.

-   **Chat History**: Saves all your conversations automatically.

-   **Dark/Light Mode**: Beautiful UI with theme switching.

-   **Responsive Design**: Works on desktop and mobile.

### ⚙️ DevOps & Backend

-   **Dockerized**: Full stack (Frontend, Backend, Database) runs in containers.

-   **PostgreSQL**: Robust relational database for user data and history.

-   **AWS Ready**: Configured for deployment on EC2.

-   **Cloudflare Tunnel**: Secure exposure without opening firewall ports (Optional).

---

## 🏗️ Architecture

```mermaid

graph TD

    User[👤 User] -->|HTTPS| Cloudflare[☁️ Cloudflare Tunnel]

    Cloudflare -->|HTTP| Frontend[🖥️ Next.js Frontend]

    Frontend -->|API Calls| Backend[⚙️ FastAPI Backend]

    

    subgraph AWS EC2

        Frontend

        Backend

        DB[(🐘 PostgreSQL)]

    end

    

    Backend -->|SQL| DB

    Backend -->|Vectors| Pinecone[(🌲 Pinecone)]

    Backend -->|Inference| Groq[(⚡ Groq LPU)]

```

## 🌐 Live Demo

The application is live and accessible via a secure Cloudflare Tunnel:

👉 **[Launch GenAI Chatbot](https://away-lucy-layer-extension.trycloudflare.com)**

*(Note: This is a demo environment running on AWS Free Tier. Initial load might take a few seconds.)*

---

## 📸 Screenshots

### Landing Page

![Landing Page](docs/screenshots/landing_hero.png)

### Tech Stack & Architecture

![Tech Stack](docs/screenshots/tech_stack.png)

### Developer Profile

![Developer Profile](docs/screenshots/developer_profile.png)

### Secure Login

![Login Page](docs/screenshots/login_page.png)

### Chat Interface

![Chat Welcome](docs/screenshots/chat_welcome.png)

### File Upload & RAG

![File Upload](docs/screenshots/file_upload.png)

### Admin Dashboard

![Admin Dashboard](docs/screenshots/admin_dashboard.png)

### User Management

![User Management](docs/screenshots/user_management.png)

---

## 🛠️ Tech Stack

| Component | Technology |

| :--- | :--- |

| **Frontend** | React 18, Next.js 14, TypeScript, Tailwind CSS |

| **Backend** | FastAPI (Python 3.11), SQLAlchemy, Pydantic |

| **Database** | PostgreSQL 15 (Data), Pinecone (Vectors) |

| **AI Model** | Llama 3-8b-8192 (via Groq API) |

| **Infrastructure** | Docker, Docker Compose, AWS EC2, Nginx/Cloudflare |

---

## 🚀 Quick Start (Local)

### Prerequisites

-   Docker & Docker Compose

-   Git

### 1. Clone the Repository

```bash

git clone https://github.com/miashraf1818/genai-rag-chatbot.git

cd genai-rag-chatbot

```

### 2. Configure Environment

Copy the example env file:

```bash

cp .env.example .env

```

Edit `.env` and add your API keys (Groq, Pinecone, Google OAuth).

### 3. Run with Docker

```bash

docker compose up -d --build

```

### 4. Access the App

-   **Frontend**: [http://localhost:3000](http://localhost:3000)

-   **Backend API**: [http://localhost:8000/docs](http://localhost:8000/docs)

---

## ☁️ Deployment (AWS EC2)

The application is successfully deployed on AWS EC2.

*   **Infrastructure**: Docker Compose on Ubuntu `t2.micro`.

*   **DNS Strategy**: `nip.io` for wildcard resolution (Google OAuth compliance).

*   **Security**: Cloudflare Tunnel for secure HTTPS access.

👉 **Read the full [Deployment Summary & Technical Notes](DEPLOYMENT_SUMMARY.md)** for architecture details, performance benchmarks, and design decisions.

### Brief Overview:

1.  Launch an Ubuntu EC2 instance.

2.  Install Docker & Git.

3.  Clone repo and set production `.env`.

4.  Run `docker compose up -d --build`.

---

## 📂 Project Structure

```

genai-rag-chatbot/

├── backend/            # FastAPI Application

│   ├── api/           # API Routes (Chat, Auth, Files)

│   ├── database/      # DB Models & Connection

│   ├── llm/           # Llama 3 & Groq Logic

│   └── vectorstore/   # Pinecone Integration

├── frontend/           # Next.js Application

│   ├── app/           # React Pages & Components

│   └── public/        # Static Assets

└── docker-compose.yml  # Container Orchestration

```

---

## 🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

## ⚖️ Trade-offs & Design Decisions

Every architecture has trade-offs. Here is how we addressed common RAG challenges:

| Challenge | Our Solution | Why? |

| :--- | :--- | :--- |

| **Latency** | **Groq LPU** | Traditional LLM inference is slow. Groq provides near-instant tokens/sec, essential for a "chat" feel. |

| **Scalability** | **Next.js + Docker** | Unlike Streamlit (prototyping), Next.js is production-ready. Docker ensures consistent deployment. |

| **Security** | **OAuth2 + RLS** | We implemented strict **Row-Level Security** logic. Users can *only* query their own vectors in Pinecone. |

| **Cost** | **Serverless / Free Tier** | Designed to run on AWS Free Tier (EC2 t2.micro) + Pinecone Starter + Groq Free Beta. |

### 🚧 Current Limitations

-   **Context Window**: Limited by Llama 3's context (8k tokens). Very large documents may lose detail in retrieval.

-   **File Types**: Currently supports PDF and TXT. OCR for scanned docs is planned for v2.

-   **Memory**: The t2.micro instance (1GB RAM) requires swap memory for build processes.

---

## 📄 License

This project is licensed under the MIT License.

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Awesome

https://github.com/miashraf1818/genai-rag-chatbot

Awesome Lists containing this project

README