I built this to help resellers (people who buy items at thrift stores/estate sales and flip them on eBay) figure out what things are worth before buying.

  You take a photo, the app identifies the item and pulls comparable sales data to estimate value. It also generates listing titles and descriptions.

  Tech stack: Next.js 16, React 19, Claude API for vision/analysis, Postgres, hosted on Vercel. iOS app is Capacitor.

  The core problem: resellers make money on information asymmetry, but researching every item manually takes forever. A vintage Pyrex bowl might be worth $5 or $50 depending on the pattern—you need to know which before you buy it.

  Just launched on iOS. Web app at https://underpricedai.com

  Happy to answer questions about the Claude integration or the business model.