EcomRadar

Overview

EcomRadar is an AI-powered research assistant that scrapes Amazon and AliExpress, analyzes trends, price bands, and pain points, then provides intelligent insights through a RAG-powered interface.

Problem

Ecommerce research is manual and time-consuming. Finding product opportunities requires scraping multiple sites, analyzing reviews, tracking prices, and synthesizing insights — a process that takes days per product category.

Solution

• Automated Scraping: Playwright-based scrapers for Amazon and AliExpress
• Data Pipeline: Product data, reviews, prices, and trends stored in Postgres
• RAG Assistant: Embeddings-based retrieval over structured product data
• Smart Analysis: AI-generated product shortlists, sourcing hints, and messaging angles

Tech Stack

• Scraping: Playwright for reliable, anti-detection web scraping
• Database: PostgreSQL with vector extensions for embeddings
• AI/RAG: OpenAI embeddings + retrieval QA over product corpus
• Backend: Supabase for database, storage, and API
• Frontend: Next.js + React for research dashboard

Impact

• 10-20x throughput: Days → hours for comprehensive product research
• Consistent briefs: Structured outputs for sourcing and marketing teams
• Data-driven decisions: Historical price trends and demand signals

Features

• Real-time product tracking across marketplaces
• Review sentiment analysis and pain point extraction
• Price band analysis and competitor comparison
• AI chat interface for natural language queries
• Export to CSV/PDF for sharing with teams