SNAVEUK
Background

Castleman Books

Castleman Books, a leading independent bookseller, approached SNAVEUK with a challenge: to streamline their online listing process for second-hand books and media across eBay.

Background

Overview

Castleman Books, a leading independent bookseller, approached SNAVEUK with a challenge: to streamline their online listing process for second-hand books and media across eBay. They wanted to minimise manual entry, reduce listing errors, and accelerate time to market for their extensive inventory.

Over two phases, SNAVEUK designed and delivered a bespoke mobile and backend solution that automates product recognition using ISBN scanning and advanced image-based AI — resulting in a faster, more intelligent way to list items

The Challenge

Castleman Books faced a time-consuming, manual listing process for thousands of second-hand books, CDs, and DVDs. Their goals included:

  • Reducing the need for manual entry of product details
  • Streamlining eBay listings through structured metadata
  • Expanding recognition capabilities to cover media without barcodes
  • Creating a mobile-first workflow suitable for warehouse and retail environments

Our Solution

We delivered the solution in two integrated phases:
Phase 1: ISBN-Based Auto-Listing System

We built a mobile application that scans ISBN barcodes and auto-generates eBay listings by retrieving structured data using the EPID (eBay Product ID). This included:

  • Barcode scanning via the mobile device’s camera
  • Automated metadata retrieval and listing population
  • Backend integration with Castleman’s existing inventory system
  • Direct submission to eBay with minimal manual input
Phase 2: AI-Powered Image Search Integration

To support non-book media and items without barcodes, we extended the system with an AI-driven image recognition feature:

  • Built-in support for photographing book covers, CDs, or DVDs
  • Integration with eBay’s Search by Image API to retrieve EPIDs from visual data
  • Backend processing pipeline to validate, enrich, and list the item
  • Intelligent handling of edge cases where no match is found

Results

  • Over 75% time savings in listing preparation
  • Higher listing accuracy and reduced errors
  • Improved scalability for handling larger inventories
  • AI-driven discovery of products not previously listed due to missing ISBNs
  • A sustainable support and maintenance agreement to keep the system running smoothly