Back to Selected Work

AI Reseller Toolkit + Chrome Extension

VintOps — multi-model AI toolkit for vintage-fashion resellers

Role: Full-Stack Developer
Dates: 2025

Overview

Resellers spend hours researching prices, writing listings, and manually posting to multiple marketplaces. VintOps compresses that into one workflow: snap a photo, get comps and pricing data, generate optimized listings for each platform, and publish — with a Chrome Extension that auto-fills the forms on Poshmark, Depop, Vinted, and others. Built end-to-end as an installable PWA with Stripe billing.

Responsibilities

  • Built reverse image search via SerpAPI and Google Lens, with similarity ranking by Gemini to find comparable listings across marketplaces
  • Developed cross-marketplace pricing intelligence combining Google Shopping and eBay sold data with summary statistics
  • Implemented listing generation with Claude that optimizes copy and structure per marketplace — one input, multiple platform-tailored outputs
  • Built an image studio on Gemini for AI-enhanced product photography with customizable backgrounds
  • Created a Manifest V3 Chrome Extension that scans marketplace listing forms and auto-fills them from generated listings via LLM field mapping
  • Implemented Clerk auth synced between the web app and Chrome Extension
  • Integrated Stripe checkout, subscription management, and customer portal with webhook-driven billing

Primary Technologies

Next.js 15React 19TypeScriptTailwindClaude (Anthropic)Gemini (Google)SerpAPI / Google LensClerk AuthStripeNeon PostgreSQLDrizzle ORMVercel BlobChrome Extension (Manifest V3)Vercel

Outcome

A working product that orchestrates multiple AI models — Claude for listing copy, Gemini for visual search and product imagery — behind a single reseller workflow, with a companion Chrome Extension and Stripe billing. Solo-built from first commit to live PWA.

The site is live at vintops.com, and a few screenshots of the app are on the landing page. Contact me to see more.