Senior Magento Data Engineer

SCOPE 1 — MAGENTO PRODUCT DATA CORRECTION & STANDARDISATION Fixed Price: $2,000 1. Existing Catalogue Data Correction * Analyse the existing Magento 2 catalogue of approximately 287,981 products. * Identify and decode Magento option IDs incorrectly stored instead of readable attribute values. * Correct encoded attribute values across the relevant product families. * Correct ABEC ratings and other affected technical attributes. * Fix approximately 1,375 SKUs containing Cyrillic characters. * Correct and normalise affected attribute sets and product data. * Write corrected values into the appropriate Magento fields. * Validate the corrected data after processing. 2. Technical Data Standardisation * Standardise product technical values and units. * Normalise measurements and SI units where applicable. * Map/classify products against ETIM and eCl@ss structures where the required data is available. * Apply relevant ISO / DIN / EN technical conventions to the product data. 3. Search & Indexing * Rebuild Elasticsearch indexes after data correction. * Rebuild/update Qdrant vector indexes. * Verify product search and indexing after the migration. 4. FR + EN Technical Content * Generate structured technical product descriptions and specification summaries from corrected product attributes. * Support both French and English. * Ensure generated content is based on actual structured product data. 5. Automated Data Pipeline * Build an automated validation and correction workflow for incoming ETL data. * Detect encoded values, invalid SKUs, inconsistent units and other known data issues. * Validate and standardise products before they become live. * Prevent the identified historical import errors from being reintroduced by future ETL batches. SCOPE 2 — AI PRODUCT & ATTRIBUTE IMAGE AUTOMATION Fixed Price: $2,000 1. Magento Image Detection * Detect products without required images. * Detect attributes and attribute options missing images. * Retrieve required data through Magento REST APIs. * Create a processing queue for image generation. 2. AI Image Generation Pipeline * Build a dedicated AI image-generation workflow. * Create an automated prompt builder based on product and attribute data. * Generate consistent technical product images. * Generate technical specification and attribute diagrams. * Maintain a consistent visual style across the catalogue. * Generate square images with a target final resolution of 2048x2048. 3. Image Processing * Remove backgrounds where required. * Upscale images to the required resolution. * Sharpen and optimise image quality. * Convert images to WebP and compress them. * Add relevant metadata such as SKU and attribute information. 4. Magento Integration * Automatically upload generated images to Magento. * Assign product images to the correct SKUs. * Store and reference attribute/specification images appropriately. * Implement the required Magento integration and display templates. 5. Full Automation * Automatically detect missing images. * Generate appropriate prompts. * Generate images through the selected AI provider. * Process and optimise generated images. * Upload and assign images to Magento. * Implement logging and error handling. * Run the workflow through scheduled/cron-based processing for future products. 6. Quality Control * Validate generated images against the required technical style. * Check resolution, orientation and image quality. * Ensure product images do not contain unwanted text or visual artifacts. * Validate attribute diagrams for consistency and clarity. 7. Documentation * Document the complete architecture and workflow. * Document configuration and maintenance procedures. * Provide documentation for future operation and maintenance. TOTAL PROJECT Scope 1 — Magento Data Correction & Standardisation: $2,000 Scope 2 — AI Product & Attribute Image Automation: $2,000 TOTAL FIXED PRICE: $4,000 The objective is to deliver both systems as reusable and automated workflows, so they can continue processing future catalogue imports rather than being only a one-time correction.