AI automation · Product data · Magento

AI product & catalogue enrichment for e-commerce

Large catalogues rarely have the product data they need: missing attributes, inconsistent naming, thin descriptions and no spec sheets. I build AI-assisted pipelines that collect manufacturer data, structure it with Claude (Anthropic), map it to a clean attribute taxonomy and push it into Magento, with every change logged.

Discuss your catalogue Magento experience
33K+SKUs in the enriched catalogue
300+brands mapped
~90%less manual processing across AI-automated workflows
Case study

Enriching a 33K-SKU tools catalogue.

Catalogue automation · Speedex Group, Dubai

AI-driven product enrichment & Magento sync

Internal admin platform · speedextools.com (Magento 2)

Context
A tools retailer with a 33,000+ SKU catalogue across 300+ brands, running on Magento 2.
Problem
Raw product data from manufacturers was incomplete and inconsistent, and the attribute taxonomy was messy, so enriching products by hand was slow.
My role
Designed and built the platform end to end, as part of a wider initiative automating cross-departmental workflows with AI.
Approach
Per-brand crawl engines collect data from brand websites; Claude structures and curates the raw data; the attribute taxonomy is normalised; branded spec-sheet PDFs are generated; attributes and products are pushed to Magento over REST. Work runs as batched jobs with backup/restore and a full audit trail.
Outcome
33K+ SKUs and 300+ brands brought into one enrichment workflow. The wider AI automation initiative cut manual processing effort by approximately 90%.
PHP 8.3Claude AIMagento RESTMySQLWeb scraping
How the pipeline works

From messy source data to a clean store.

  1. CollectPer-brand crawl engines gather product information from manufacturer websites, so each brand's site structure is handled on its own terms.
  2. Structure with AIClaude turns raw, unstructured product content into structured fields and curated descriptions.
  3. Map the taxonomyAttributes are normalised into one consistent taxonomy, so the same property isn't stored five different ways.
  4. Generate assetsBranded spec-sheet PDFs are produced from the structured data.
  5. Sync to MagentoAttributes and products are pushed to Magento over the REST API in batched jobs.
  6. Keep it safeBackup/restore and a full audit trail mean any change can be traced and rolled back.
Where it fits

Useful when your catalogue outgrows manual work.

Large or growing catalogues

Thousands of SKUs across many brands, where hand-writing attributes and descriptions can't keep up.

Inconsistent attributes

Filters and comparisons break because the same property is named and formatted differently across products.

Magento and ERP stores

Enriched data has to land in Magento and stay consistent with the ERP. I also build ERP-to-Magento catalogue, pricing and order sync.

Have a catalogue that needs enriching?

I'm based in Dubai and open to roles and focused consulting on AI product-data automation, catalogue enrichment and Magento integration.

Get in touch LinkedIn
Magento lead developer →ERP integration, payment APIs, multi-site performance and replatforming. Full resume →Roles, certifications (PMP, PSM I) and education (MBA).