Car dealerships deal with vehicle catalogs from dozens of providers, each with their own PDF format, layout, and structure. Manually extracting specs, pricing, and discounts from those documents is tedious and error-prone.

A car dealership reached out to me to help fix this specific issue.

It ingests PDFs regardless of format, uses AI extraction to pull out full vehicle data, and loads everything into clean structured tables. The team can then review, edit, add missing entries, and export to Excel in a few clicks. Every file keeps a full extraction detail view so nothing gets lost or misattributed.

Built with Next.js, OpenAI, and AWS. Genuinely one of those projects where you can feel the immediate impact the moment it goes live.