Blinkit Clone

A fully functional clone of the Blinkit website, developed as a project to gain hands-on experience with various technologies.

Tech Stack Used

• Frontend: HTML, CSS, JavaScript, React.js, Next.js, Tailwind CSS, DaisyUI
• Backend & Database: PostgreSQL, Redis
• State Management: Redux
• Tools Used: Cursor, Claude, ChatGPT, Docke

Why this Project?

This project was undertaken to:

• Gain a deeper understanding of real-time inventory management and geospatial queries
• Implement caching mechanisms for handling large-scale inventory efficiently
• Work with PostgreSQL and Redis to optimize performance
• Learn about store assignment based on geolocation

Development Process

• Prototype & Planning: Documented the initial prototype and database schema in Notion.
• Database Design: Created normalized tables and integrated PostgreSQL with PostGIS for geospatial queries.
• Caching Optimization: Implemented Redis for frequently accessed inventory data.
• Backend Implementation: Built APIs for store assignment, cart management, and order processing.
• Frontend Development: Used Next.js with Tailwind CSS for a responsive UI.

Challenges

Database Structure & Choice

• Primary Database: PostgreSQL (with JSONB for semi-structured data)
• Geospatial Queries: PostGIS extension to store and query locations
• Caching Layer: Redis to improve query performance and reduce database load
• Data Flow: PostgreSQL acts as the source of truth, while Redis caches store inventory for faster access.

Why PostgreSQL?

• Supports relational data modeling
• Ensures ACID compliance for transaction safety
• Scalability
• Provides JSONB support for flexible data storage
• Offers efficient indexing (e.g., on sub_category, brand_name for fast queries)

Tables made

• user
• Categories
• Sub_categories
• Products
• stores
• orders

Assigning the Nearest Store to a User

• Fetch the user's coordinates.
• Identify all stores within a 5km radius using PostGIS geospatial queries.
• among those stores the distance is calculated and the nearest one is assigned to the user.

Handling Vast Inventory Data Efficiently

• Directly querying PostgreSQL for each request was inefficient.
• Implemented a Redis caching layer for store inventories.
• Redis stores real-time inventory updates and uses TTL for cart items.
• Inventory updates occur on cart addition (with TTL expiration) and checkout.

Demo Video

demo-video.mp4