Ecommerce API RAG DRF

Ecommerce-API-RAG-DRF is a robust backend system for eCommerce that integrates an intelligent virtual assistant (RAG Chatbot) capable of answering real-time inventory questions using semantic search and Large Language Models (LLMs).

Description

Final project for the "API Development with Microservices Architecture" course. It demonstrates proficiency in Python, Django REST Framework, and AI model integration (LLM) via RAG (Retrieval-Augmented Generation). The chatbot Techty utilizes semantic search over a vector database to provide contextual responses about store products.

Key Features

• Full REST API with JWT authentication
• RAG Chatbot featuring semantic product search
• Real-time currency conversion (USD -> VES)
• Shopping cart and order management system
• Interactive documentation with Swagger/Redoc
• Rate limiting and per-endpoint throttling

Architecture

graph TD
    A[Root] --> B[core]
    A --> C[productos]
    A --> D[tienda]
    A --> E[usuarios]
    A --> F[chatbot]

    B --> B1[settings.py]
    B --> B2[urls.py]

    C --> C1[Product/Category Management]
    D --> D1[Cart and Purchase Orders]
    E --> E1[JWT User Management]
    F --> F1[AI Logic and RAG Vectors]

    F --> F2[Groq API / Llama.cpp]
    F --> F3[Upstash Vector DB]

Loading

• core/: Central configuration for the Django project
• productos/: Models and views for inventory, including currency conversion logic
• tienda/: Management of the purchasing workflow (Cart -> Order -> Stock)
• usuarios/: Custom authentication using SimpleJWT
• chatbot/: Integration with Groq/Llama.cpp and Upstash Vector for intelligent assistance

Prerequisites

• Python: 3.10 or higher
• pip: Latest version
• Venv: Recommended for dependency isolation

Optional:

• PostgreSQL (if you prefer to use it instead of SQLite)
• Groq Account (free, for the LLM)
• Upstash Account (free, for the vector database)

Installation

1. Clone the repository

git clone https://github.com/DaniDevGS/Ecommerce-API-RAG-DRF.git
cd Ecommerce-API-RAG-DRF

2. Create and activate virtual environment

python -m venv venv

# Windows
venv\Scripts\activate

# Linux/Mac
source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables

Copy the example file and update it with your credentials:

copy .env.example .env     # Windows
cp .env.example .env        # Linux/Mac

Open .env and set at least the minimum required variables:

SECRET_KEY=django-insecure-change-this-key
DEBUG=True

5. Run migrations

python manage.py migrate

6. Create superuser (optional)

python manage.py createsuperuser

7. Start the server

python manage.py runserver

The server will be available at http://127.0.0.1:8000/

Environment Variables

The project works with a minimal configuration (only SECRET_KEY and DEBUG). Additional variables enable extra features.

Required

Variable	Description	Example
SECRET_KEY	Django secret key	django-insecure-your-key
DEBUG	Development mode	True

LLM (Choose your preferred provider)

Variable	Description	Where to get it
GROQ_API_KEY	Groq API Key (free)	console.groq.com
GROQ_API	Groq Endpoint	https://api.groq.com/openai/
LLAMACPP_API	Local Llama.cpp Endpoint	http://localhost:8080/

Vector Database

Variable	Description	Where to get it
UPSTASH_VECTOR_REST_URL	Upstash Vector URL	console.upstash.com
UPSTASH_VECTOR_REST_TOKEN	Upstash Vector Token	console.upstash.com

SQL Database (Optional)

Variable	Description	Example
DATABASE_URL	PostgreSQL connection URL	postgresql://user:pass@host:5432/db

Minimal vs. Full Setup

MINIMAL - REST API Only

SECRET_KEY=your_key_here
DEBUG=True

WITH LLM - Chatbot responds without product context

SECRET_KEY=your_key_here
DEBUG=True
GROQ_API_KEY=gsk_your_key
GROQ_API=https://api.groq.com/openai/

FULL SETUP - Chatbot with semantic product search & REST API functionality

SECRET_KEY=your_key_here
DEBUG=True
GROQ_API_KEY=gsk_your_key
GROQ_API=https://api.groq.com/openai/
UPSTASH_VECTOR_REST_URL=https://your-vector.upstash.io
UPSTASH_VECTOR_REST_TOKEN=your_token

API Endpoints

Method	Endpoint	Auth	Description
GET	/	No	Swagger UI (interactive documentation)
POST	/api/auth/login/	No	Login and obtain JWT
POST	/api/auth/refresh/	No	Refresh JWT token
POST	/api/usuario/	No	Register user
GET	/api/usuario/	Admin	List users
GET	/api/productos/	No	List/filter products
GET	/api/productos/<id>/	No	Product details
GET	/api/categorias/	No	List categories
GET	/api/subcategorias/	No	List subcategories
GET	/api/carritos/	JWT	View my cart
POST	/api/items-carrito/	JWT	Add item to cart
PATCH	/api/items-carrito/<id>/	JWT	Edit cart item
DELETE	/api/items-carrito/<id>/	JWT	Delete cart item
POST	/api/ordenes-compra/	JWT	Create order from cart
GET	/api/ordenes-compra/	JWT	List my orders
POST	/api/chatbot/	JWT	Query chatbot (RAG)
GET	/api/chatbot/	JWT	View chat history

Example: Login

curl -X POST http://127.0.0.1:8000/api/auth/login/ \
  -H "Content-Type: application/json" \
  -d '{"username": "your_username", "password": "your_password"}'

Example: Querying the Chatbot

curl -X POST http://127.0.0.1:8000/api/chatbot/ \
  -H "Authorization: Bearer your_jwt_token" \
  -H "Content-Type: application/json" \
  -d '{"pregunta": "What laptops do you have available?"}'

Project Structure

backend/
├── core/                # Django Configuration (settings, urls, wsgi)
├── productos/          # Models: Category, Subcategory, Product
├── tienda/              # Models: Cart, CartItem, PurchaseOrder
├── usuarios/            # User Management with JWT
├── chatbot/             # RAG: rag.py (LLM), vectors.py (Upstash)
├── media/               # Uploaded media files
├── requirements.txt    # Python Dependencies
├── .env_example        # Environment variables template
└── manage.py            # Django management commands

Technologies

Component	Technology	Version
Framework	Django	5.2
REST API	Django REST Framework	3.16

License

This project is licensed under Personal Use Only. Distribution, sale, or usage in production environments is strictly prohibited. Refer to the LICENSE file for details.