Product-Price Comparison AI

I want to bring to life an AI-driven browsing assistant that automatically visits online stores, extracts the key details of physical products, and presents a clear side-by-side comparison of price, shipping cost, ratings, and core specifications. The first release should focus strictly on standard e-commerce sites (Amazon, eBay, Walmart, etc.), but the architecture needs to stay modular so we can later plug in travel deals or local-service providers with minimal refactoring.

Feel free to choose the stack you are most comfortable with; Python + Selenium/Playwright, Node + Puppeteer, or an equivalent headless-browser approach is fine, provided it can handle dynamic pages, lazy-loaded content, and the usual anti-scraping countermeasures. A lightweight model that handles product-attribute normalization and synonym matching will be valuable, but I am open to classical rule-based parsing if it proves faster and equally accurate.

There is no hard deadline, yet steady milestones matter. I would like to see a quick proof of concept on two major retailers first, then an incremental rollout to five more. Clean, well-commented code, unit tests, and concise setup notes are non-negotiable so that I can host the service on my own VPS when we finish.

Deliverables
• Working prototype covering at least two large e-commerce sites
• Scalable scraper/agent code with documented modules for additional sites
• Comparison engine that outputs a unified JSON and a simple web view
• Read-me style deployment guide and short video walkthrough

For acceptance, the tool must detect identical SKUs across the chosen stores, report the cheapest total cost (item + shipping), and update results within 60 seconds of a user query. Let me know which libraries you plan to lean on and any API keys I should secure ahead of time.