Project Snapshot

Business need: Reduce repetitive manual work and data-quality risk across multi-platform product preparation and monthly inventory reconciliation.

My role: Designed and operated two repeatable, AI-assisted workflows for product-data preparation and inventory reconciliation. I defined the data mappings, validation rules and exception handling, and used Codex to support repetitive transformation and processing.

Scope: 262 SKUs · Shopify · Shopee · Lazada · TikTok Shop · product data · inventory reconciliation · validation · exception handling

Workflow 1 — Multi-Platform Product Data

The product-data workflow transforms source records into platform-ready outputs for Shopify, Shopee, Lazada and TikTok Shop.

• Standardised SKU, barcode/EAN, price, dimensions, weight, category, variant and inventory fields
• Mapped source data into platform-specific upload structures
• Prepared titles, descriptions, specifications and required marketplace fields
• Added checks for missing, duplicate, malformed or conflicting values
• Flagged ambiguous records instead of guessing
• Kept pricing, identifiers, variants, stock and shipping details under manual review before upload
• Validated published listings after implementation

Result

For a typical 30–50 SKU batch, estimated preparation time was reduced from up to 2 workdays to approximately 1–3 hours, with validation and final review built into the workflow.

Workflow 2 — Inventory Reconciliation

The monthly reconciliation workflow consolidates consignment, sales, delivery-order, balance-master and platform records into a repeatable review process.

• Standardised SKU / Art No. values before matching
• Consolidated multiple operational files
• Applied sales deductions and delivery-order additions