20 Minutes → 3 Seconds: Automated Product Listings for an Automotive Parts Marketplace
The Challenge
An automotive parts marketplace managing a catalog of approximately 500,000 SKUs sold across Amazon, eBay, and Shopify was drowning in manual data work.
Each product required roughly 20 minutes of human effort to transform raw manufacturer data — often riddled with unstructured codes and missing product names — into polished, multi-channel listings. The full data cleansing pipeline ran on a 9-day processing cycle, creating a bottleneck that limited how quickly inventory could get to market.
What We Built
60x rebuilt the entire data pipeline and content generation system. Titles, descriptions, bullet points, meta tags, and marketplace categorisations now generate simultaneously across all three sales channels.
A multi-agent enrichment system resolves ambiguous data automatically — identifying products from incomplete manufacturer codes, normalising specifications, and generating channel-specific content — with an estimated 97% of items requiring zero human review.
The Results
- Per-SKU processing: from ~20 minutes of manual effort to 1–3 seconds, automated
- Full pipeline cycle: compressed from 9 days to a few hours
- Accuracy: previously title-only coverage at 95%; now full content across 3 channels with 97% requiring no human review
The Compounding Effect
Compressing the data pipeline from 9 days to a few hours didn't just free up staff time — it removed the bottleneck that was limiting how quickly 500,000 SKUs could get to market across three sales channels.
The immediate time savings were real, but the strategic value was in unlocking capacity that didn't exist before. The business can now onboard new product lines and respond to market changes at a pace that was previously impossible.