TechMart - AI Powered Electronics E-Commerce Platform

TechMart is a full-stack Electronics E-Commerce platform built using Spring Boot 3, React 19, MySQL, Elasticsearch, Redis, and AI.

The project demonstrates modern backend development with secure authentication, caching, search relevance tuning, automated email updates, secure Razorpay payments, and an intelligent AI shopping assistant.

---

Features

Authentication & Security

• JWT Authentication: Secure stateless authentication using JWT tokens.
• HttpOnly Cookie Authentication: Tokens are stored in secure, HttpOnly cookies for enhanced protection against XSS attacks.
• Role-Based Authorization: Restricts access to endpoints based on roles: ROLE_CUSTOMER and ROLE_ADMIN.
• Spring Security Configuration: Custom filters and security rules governing public and protected routes.

---

Product Management (Admin Only)

• CRUD Operations: Admins can add, update, delete, and view products.
• Bulk Import: Endpoint supporting bulk insertion of product data.
• Product Categories: Dynamic management of category relationships.
• Cloudinary Image Upload: Automatic uploading of product images to Cloudinary from image URLs.

---

Shopping & Ordering (Customer)

• Shopping Cart: Add items, update quantities, and remove products dynamically.
• Wishlist: Save favorite products for later purchase.
• Profile & Address Management: Add, view, edit, and delete multiple shipping addresses.
• Secure Checkout & Buy Now: Guided checkout flow with address selection and payment verification.
• Payment Gateway Integration: Embedded Razorpay integration for secure payments.
• Order History: View complete personal order details, statuses, and payment states.

---

Order & Inventory Management (Admin Only)

• Global Order Overview: View all orders placed across the system.
• Order Status Updates: Change order states (PLACED, PROCESSING, SHIPPED, DELIVERED, CANCELLED).
• Automated Inventory Stock Handling:
  • Stock is automatically decremented upon a successful purchase.
  • Stock is automatically restored (restocked) if an order is cancelled by the Admin.

---

Automated Email Notifications

• Spring Mail Service: Automatically sends a detailed HTML/text order confirmation email (via Gmail SMTP) upon successful payment verification.
• Includes details like Order ID, products, quantity, total amount, and delivery message.

---

Reviews & Ratings

• Customer Reviews: Write text feedback and rate products from 1 to 5 stars.
• Dynamic Aggregate Metrics: Automatically calculates average ratings and counts reviews per product.

---

Smart Search

Powered by Elasticsearch and optimized with Redis Caching:

• Full Text Search: Queries matching name (boosted ^3), category (boosted ^2), and description.
• Typo Tolerance: Enabled using Elasticsearch fuzzy search query configuration ("fuzziness": "AUTO").
• Advanced Relevance & Popularity Score: Custom function_score query combining text relevance with a calculated salesScore (popularity factor based on ratings, sales volume, stock availability, review count, and product freshness).
• Search Autocomplete suggestions: Quick suggestions starting with the query prefix.
• Redis Cache Layer: Caches search queries, autocomplete suggestions, product lists, and details. Automatically evicts relevant cache namespaces when products are added, modified, or deleted.

---

AI Shopping Assistant

Powered by Groq API (llama-3.3-70b-versatile model):

• Intent Extraction: Parses user queries using structured JSON format into categories, brand, budget constraints, keywords, purpose, and sort criteria.
• Filtered Product Feeds: Before sending to the LLM, the system queries Elasticsearch and filters products programmatically by category, brand, budget range, and stock availability.
• Context-Aware Conversations: Stores conversation history in MySQL to handle follow-up queries (e.g., "cheaper options", "what about Dell?", "battery life?").
• Formatted Recommendations: Returns clean, beautifully structured markdown templates displaying rank, name, product ID, pricing, ratings, review count, total sold, and detailed reasoning.
• Chat Management: Supports creating, switching, and deleting multiple chat threads.

---

Tech Stack

Backend

• Java 17
• Spring Boot 3.5.3
• Spring Security
• Spring Data JPA & Hibernate
• Spring Data Elasticsearch
• Spring Data Redis
• Spring Boot Starter Mail
• Lombok
• Maven

Frontend

• React 19
• Vite
• Axios
• React Router DOM 7
• React Icons
• Vanilla CSS (CSS Modules)

Database & Third-Party Services

• MySQL (Relational storage)
• Elasticsearch (Search engine)
• Redis (Caching server)
• Groq API (Llama 3.3 70B LLM)
• Cloudinary (Image hosting)
• Razorpay (Payment gateway)

---

Architecture

React (Vite)
      │
      ▼ (Axios + HttpOnly JWT Cookie)
Spring Boot REST API
      │
      ├──────────────► MySQL
      │
      ├──────────────► Elasticsearch
      │
      ├──────────────► Redis
      │
      ├──────────────► Cloudinary
      │
      ├──────────────► Razorpay API
      │
      ├──────────────► SMTP Mail (Gmail)
      │
      └──────────────► Groq AI (Llama 3.3)

---

Project Structure

Backend Package Structure

src/main/java/com/example/TechMart/
 ├── auth           # Register, Login, Current User, Jwt cookies
 ├── user           # Users entity & repository
 ├── category       # Category controller, entity & repository
 ├── product        # Product controller, image uploads, services
 ├── review         # Review and Rating controller, entity & repository
 ├── wishlist       # Wishlist and Wishlist Item management
 ├── cart           # Cart and Cart Item management
 ├── order          # Order, Order Items, Payments & Email services
 ├── elasticsearch  # Search services, indexing documents, repositories
 ├── ai             # Chat, Conversations, Intent, Groq API services
 ├── profile        # Shipping Address entities, controllers & services
 ├── security       # JWT Filter, security filters & CORS settings
 └── config         # Third-party credentials & Bean configurations

---

AI Workflow

User Query
      │
      ▼
Intent Extraction (Groq JSON Mode)
      │
      ▼
Elasticsearch Search (Function Score Query)
      │
      ▼
Category, Brand, Budget, & Stock Filters (Spring Boot Service)
      │
      ▼
Popularity & Freshness Ranking (Based on Sales Score)
      │
      ▼
Groq AI Recommendation (llama-3.3-70b Context)
      │
      ▼
Response Saved to MySQL Chat History

---

Database Schema

Main Tables

• Users: Authentication data, emails, and roles (ADMIN / CUSTOMER).
• Products: Electronic items details, prices, stocks, sales counters, and ratings.
• Categories: Parent groupings for products (e.g., Laptops, Mobiles, Tablets).
• Address: Shipping address info linked to Users and Orders.
• Reviews: Customer text reviews and star ratings.
• Cart & Cart Items: Temporary customer shopping cart records.
• Wishlist & Wishlist Items: Customer marked favorites.
• Orders & Order Items: Purchased records, shipping destinations, and quantities.
• Payment: Payment transaction records, Razorpay IDs, and transaction status (CREATED, SUCCESS, FAILED).
• Conversation: Grouping records for AI chat histories.
• Chat Messages: Chat exchanges containing messages and roles (USER / ASSISTANT).

---

Roles & Access Control

Customer

• Register & Login
• Browse, detail-view, and search products
• Maintain Cart & Wishlist
• Add & edit shipping addresses
• Order products via secure checkout & verify payment
• Review and rate products
• Create, manage, and hold chats with the AI Shopping Assistant
• View order history & transaction logs

Admin

• Complete Product CRUD (Add, Edit, Delete, Bulk import)
• Category Management
• Global Order Management (view all orders and change statuses)
• Stock monitoring & inventory management

---

Setup & Configuration

Prerequisites

Make sure you have the following installed:

• Java 17 SDK
• Maven
• MySQL Server
• Elasticsearch (running on http://localhost:9200)
• Redis Server (running on port 6379)
• Node.js & npm

1. Database & Service Setup

Create a MySQL database named techmart:

CREATE DATABASE techmart;

2. Configure Backend Applications

Modify TechMart_Backend/src/main/resources/application.properties with your configurations:

spring.datasource.url=jdbc:mysql://localhost:3306/techmart
spring.datasource.username=YOUR_MYSQL_USERNAME
spring.datasource.password=YOUR_MYSQL_PASSWORD

# Cloudinary
cloudinary.cloud-name=YOUR_CLOUDINARY_NAME
cloudinary.api-key=YOUR_CLOUDINARY_API_KEY
cloudinary.api-secret=YOUR_CLOUDINARY_API_SECRET

# Razorpay
razorpay.key.id=YOUR_RAZORPAY_KEY_ID
razorpay.key.secret=YOUR_RAZORPAY_KEY_SECRET

# Elasticsearch
spring.elasticsearch.uris=http://localhost:9200

# Redis
spring.data.redis.host=localhost
spring.data.redis.port=6379

# SMTP Email (Gmail Example)
spring.mail.host=smtp.gmail.com
spring.mail.port=587
spring.mail.username=YOUR_EMAIL@gmail.com
spring.mail.password=YOUR_GMAIL_APP_PASSWORD
spring.mail.properties.mail.smtp.auth=true
spring.mail.properties.mail.smtp.starttls.enable=true

# Groq AI Key
groq.api.key=YOUR_GROQ_API_KEY

Note

On first startup, the application automatically creates a default administrator account if one does not already exist.

3. Run Backend

cd TechMart_Backend
mvn clean spring-boot:run

4. Configure & Run Frontend

Create a .env file under TechMart_Frontend/ (or update it):

VITE_RAZORPAY_KEY_ID=YOUR_RAZORPAY_KEY_ID

Install dependencies and run the development server:

cd TechMart_Frontend
npm install
npm run dev

Open your browser at http://localhost:5173.

---

Application Screenshots

Home Page

Login & Register

Product Details

Cart

Wishlist

Product Management

Order Management

Admin Dashboard

Elasticsearch Search

Razorpay Payment

Email Notification

AI Assistant

---

Future Improvements

• Microservices Architecture: Transition backend services to Spring Cloud.
• Dockerization: Create Docker containers for the Backend, Frontend, MySQL, Redis, and Elasticsearch.
• Kubernetes: Orchestrate containers using K8s clusters.
• Event Streaming: Implement Apache Kafka for order processing events.
• AI Recommendation Engine: Advanced user recommendation engine based on collaborative filtering.
• Voice & Image Search: Support searching products via voice commands or uploaded product images.

---

Author

Prashanth

Full Stack Developer | Spring Boot | React | Java | Elasticsearch | Redis | AI Integrations