Modern e-commerce applications demand high performance, scalability, real-time operations, secure APIs, and efficient caching. I built a complete multi-category e-commerce backend that supports Pharma, Clothing, and Restaurant modules using TypeScript, Node.js, Express.js, PostgreSQL, Redis, and AWS S3. This blog explains how I designed the backend architecture, handled data flow, integrated caching, structured APIs, and optimized the overall performance for real-world production use.

1. Choosing the Right Tech Stack

The technology stack was selected to provide an ideal balance of speed, scalability, and maintainability. TypeScript ensured cleaner code, stronger type safety, and reusable interfaces. Node.js with Express.js offered an event-driven backend capable of handling high traffic efficiently. PostgreSQL served as an ACID-compliant database suitable for structured product, user, and order data. Redis functioned as a powerful caching layer to significantly reduce response times. AWS S3 was used for reliable and scalable image storage for all product-related assets.

2. Backend Architecture: Modular & Clean

The backend follows a layered architectural pattern to maintain scalability and clarity. Controllers manage incoming requests, services contain the business logic, and repositories handle all database operations through TypeORM. Middlewares are used for authentication, request validation, and logging. This separation ensures easy debugging, better structure, and long-term maintainability.

3. PostgreSQL Database Design

The core of any e-commerce system is a well-designed relational database. I created structured tables for users, products, Pharma-specific items, clothing variants, restaurant food items, orders, carts, inventory, vendors, and categories. Each table was optimized using indexes to enable fast search capabilities and efficient filtering, which is crucial for smooth user experience in large product catalogs.

4. Redis Caching for High Performance

Redis was integrated to improve the performance of frequently accessed operations. Product listings, category filters, stock availability data, cart information, and order sessions were cached to avoid repeated database hits. This resulted in a major performance boost, reducing API response times from around 300ms down to a stable 20–25ms, significantly enhancing the user experience across all modules.

5. Handling Images with AWS S3

E-commerce platforms rely heavily on high-quality product images. AWS S3 was used to manage secure and scalable image storage. I implemented a secure upload system, generated signed URLs for controlled image access, and added compression workflows to optimize loading speed. This covered product galleries, category banners, and all media assets used across the platform.

6. Core Modules Developed

Authentication: JWT + role-based access.

Product Module: Variants, filters, categories, search.

Cart & Checkout: Dynamic price updates, coupon system.

Orders: Payment tracking, order lifecycle, vendor allocation.

Pharma Module: Dosage, expiry, medical tags.

Restaurant Module: Menu, add-ons, delivery charge logic.

Clothing Module: Sizes, colors, inventory per variant.

7. Security & Best Practices

• JWT authentication

• Password hashing (bcrypt)

• API rate limiting

• SQL injection protection

• Data validation using Zod

• Structured error handling

Conclusion

Building this multi-category e-commerce backend enhanced my understanding of scalable architecture, distributed caching, API design, and performance optimization. By leveraging TypeScript, Node.js, PostgreSQL, Redis, and AWS S3, the system is secure, efficient, and ready to scale for large production environments.

Written by Moin Baig — Backend Developer, Purple Sky Infotech