E-Commerce Manager | AI & Automation Focus
Big Box Art - Order Processing & Image Automation System



Project Overview
A comprehensive desktop application I engineered from scratch for my print-on-demand business, Big Box Art, leveraging AI-assisted development and Python programming to eliminate manual workflows and achieve end-to-end order fulfillment automation. Processing hundreds of daily orders across multiple sales channels (eBay, Amazon, Amazon Prime, Wayfair) while coordinating with various print suppliers presented significant operational bottlenecks that I resolved through custom software development.
Utilizing AI collaboration and my Python expertise, I designed and implemented sophisticated data processing pipelines, intelligent caching mechanisms, and seamless external system integrations. The application ingests purchase order data from Order Desk, processes multi-channel orders, and generates supplier-ready files with embedded product images. This solution reduced order processing time from hours to minutes, eliminated data entry errors, and enabled business scaling without proportional labor cost increases—transforming a manual, time-intensive workflow into a single-click automated process.
Business
Impact
-
Reduced order processing time from hours to minutes
-
Eliminated manual data entry errors across multiple platforms
-
Automated image preparation for multiple print suppliers
-
Streamlined shipping label generation for hundreds of daily orders
-
Consolidated multi-item orders for accurate weight calculation and cost optimization
-
Enabled scalable operations without additional staff overhead
Key
Features
Intelligent Order Processing
-
Purchase order data import with automatic field mapping and validation
-
Automated CSV generation for multiple print suppliers and product categories
-
Smart sales channel detection with conditional logic for platform-specific requirements
-
Dynamic quantity expansion and print naming with multi-copy tracking
-
Multi-purchase discount tagging for eBay bulk orders (3+ items)
Image Automation Pipeline
-
Automated image downloading from Azure Blob Storage with SAS token authentication
-
Integration with Adobe Photoshop via AppleScript for batch processing
-
Dynamic filename generation with order tracking, sales channel codes, and print specifications
-
Automatic prefix removal post-processing for clean file organization
Shipping & Fulfillment Integration
-
Automated shipping label generation with weight calculation and address consolidation
-
Tracking number converter with one-click clipboard functionality for bulk shipment updates
-
Excel export with embedded product thumbnails using PIL image processing
-
Conditional address logic (label-provided vs. full address) based on fulfillment channel
Advanced Product Handling
-
Platform-specific SKU lookup system with multi-file caching architecture
-
Real-time postcode validation and data integrity checks
-
Support for 5 product categories across 4 sales channels
-
Intelligent supplier routing based on product type and order specifications
Technical
Stack
-
Development: Claude Sonnet 4.5 | ChatGPT o1
-
Language: Python 3.12
-
GUI Framework: PyQt6 with custom dark theme and tabbed interface
-
Data Processing: Pandas for CSV manipulation and data transformation
-
Image Processing: Pillow (PIL) for thumbnail generation
-
Excel Generation: openpyxl for embedded image support
-
External Integration: AppleScript automation for Photoshop
-
Cloud Storage: Azure Blob Storage with SAS token authentication
Technical
Highlights
-
Caching System: Intelligent multi-level caching for lookup files and product data to minimize I/O operations
-
Error Handling: Comprehensive logging with detailed error tracking and CSV export for failed operations
-
Memory Management: Temporary file lifecycle management for thumbnail processing
-
UI/UX: Progress tracking with real-time status updates and visual feedback
-
Data Validation: Weight calculation from lookup tables, SKU pattern matching, and address formatting
Deployment
-
Packaged using PyInstaller for standalone macOS application
-
Resource bundling for CSV lookup files and branding assets
-
Version-controlled configuration for supplier specifications
-
Dynamic Photoshop version detection for system compatibility
Technologies: Python | PyQt6 | Pandas | Pillow | openpyxl | AppleScript | Azure Storage | Claude 4.5 | ChatGPT 5
Type: Desktop Application | Business Automation | E-commerce Integration