Technical Deep Dives

Engineering Case Studies

A look under the hood at the architectural challenges we've solved. From dynamic ERP modules to asynchronous cloud management.

1. AI-Powered ERP System (ERPModules)

Python PySide6 PostgreSQL

The "Why"

A comprehensive, modular Enterprise Resource Planning system designed to manage inventory, procurement, and financials.

Key Technical Challenges Solved

Dynamic Architecture

Implemented a plugin-style architecture where business modules (Inventory, Procurement, HR) are loaded dynamically at runtime using importlib and introspection.

State Management

Handled complex data interdependencies (e.g., receiving a PO automatically updates Inventory stock and General Ledger entries) using raw SQL transactions to ensure data integrity.

2. Smart Screen Recorder

mss numpy Pillow (PIL)

The "Why"

A high-performance desktop utility for recording screen activity and exporting it as optimized GIFs or image sequences.

Key Technical Challenges Solved

Real-time Performance

Achieved low-latency screen capture using mss and numpy for frame processing, running on a separate thread to prevent UI freezing.

Custom Encoding Pipeline

Built a custom GIF encoding engine using PIL that handles palette quantization and transparency masks manually to optimize file size and quality.

3. License Server Commander

Google Cloud SDK Subprocess Threading

The "Why"

A remote management console for a cloud-hosted license server running on Google Cloud Platform.

Key Technical Challenges Solved

Asynchronous Remote Execution

Created a threaded execution engine to run blocking SSH commands (via gcloud CLI) without freezing the GUI, streaming stdout/stderr back to a real-time console log in the UI.

Log Parsing

Implemented regex-based parsing of remote server logs to convert server timestamps to local time zones on the fly.