AI Price Arbitrage Tool -- 2

I want to build an AI-driven arbitrage engine that focuses strictly on price comparison. The goal is to give businesses a single interface that scans multiple marketplaces, wholesaler feeds, and service-provider listings, then highlights where buying low and selling high is immediately possible. The platform must: • Compare prices across three categories: physical products, digital products, and services. • Deliver real-time or near-real-time results, ranked by potential margin. • Present findings in a clean dashboard and expose the same data through a documented REST or GraphQL API so it can feed existing ERP or e-commerce systems. Core expectations 1. Data acquisition: you decide the best combination of web scraping, official APIs, or partner feeds in compliance with each site’s ToS. 2. Intelligence layer: use machine-learning or rules-based logic—whichever proves faster to market—to normalise SKUs/titles, detect duplicates, and calculate actionable spreads. 3. Business-ready output: reports downloadable as CSV/JSON and visible charts that clearly flag the arbitrage opportunity, including fees and shipping where relevant. 4. Security and scalability: deploy on AWS, GCP, or Azure with basic auth + role management so multiple teams inside a company can use it safely. Acceptance criteria • A demo account comparing at least 15 sources per category shows working results. • API returns structured data with <500 ms latency for a sample query. • Documentation and a quick-start guide are included so an in-house dev can integrate it within a day.