🧬 ProteinDB

The Problem

India's protein supplement market has hundreds of products spread across multiple e-commerce platforms. Comparing them manually across price, protein content per serving, serving size, and quality certifications is tedious and prone to error. Having said that, the bigger concern is that most "review" websites ranking these products are affiliate-driven, which means their recommendations are influenced by commission rates, not by actual product value.

What I Built

ProteinDB automates the comparison process. It scrapes product data from major Indian e-commerce platforms, normalizes inconsistent listings into a common format, and scores each product on objective value metrics:

• Protein per serving (grams) and protein per rupee (the value score that actually matters)
• Price tracking across multiple sellers to identify the best deal at any given time
• Quality indicators including certification checks where available
• A clean, filterable comparison interface where you can sort, search, and filter by what matters to you
• Zero affiliate links. The rankings are purely data-driven, and I have no financial relationship with any brand listed.

Architecture

• Scraper (Python) — Collects and normalizes product data from e-commerce sites
• Frontend (Next.js + TypeScript) — Server-rendered comparison UI with search and filters
• Data Pipeline (GitHub Actions) — Automated scraping and data refresh
• Static Export — Deployed as static site on GitHub Pages

Tech Stack

Links