Summary
- Leveraged OCR (Optical Character Recognition) to automate product base management, replacing days of manual entry with a 2-minute automated batch process.
- Engineered a lightweight, no dll dependency utility with a UI optimized for a multi-generational workforce, requiring zero formal training for elderly employees.
- Developed a modular architecture supporting 20+ distinct business scenarios, including automated image concatenation, notification templating, and cloud data synchronization.
PZPToolkit is a high-impact standalone utility developed for PZP Wuhan to formalize complex workflows and eliminate data entry bottlenecks.
The project’s centerpiece is an intelligent OCR (using PaddleOCR) and Image Processing engine that transformed a multi-day manual watermarking and tagging process into a fully automated pipeline.
By leveraging text recognition to identify product tags and apply context-aware watermarks, the system processes entire product batches in under 2 minutes, whilst the task previously required days of manual labor.
To ensure 100% data integrity, the application utilizes a human-in-the-loop validation system, automatically flagging and interrupting only when OCR confidence is low or required fields are missing.
Apart from OCR, this specialized toolset includes over 20 distinct functions: ranging from daily report generation and spreadsheet refining to cross-department data synchronization.
The data is stored in a fixed location, fetching and updating when required (to ensure flexiblity and timeliness). Python scripts & environment will be installed automatically within the single .exe application environment designed for maximum accessibility across all employee age groups.
