Raezan Cabang

I've built AI-assisted prescription processing software at Cubeworks Technology Consulting and Solutions, combining Python, React, REST APIs, OCR, and Ollama to improve handwritten prescription interpretation accuracy by 30%.

I trained and optimized an OCR model on approximately 4,000 handwritten prescription images using Google Colab and CUDA, reducing training time by 25% while improving recognition of complex medical terms.

I've also automated ServiceNow ticket reporting with Excel VBA, replacing manual VLOOKUP-based updates with a one-click workflow that preserves historical records, prevents duplicates, and adapts to changing columns.

Through projects including Hunt Buddy, an offline-capable Learning Management System, and a Java inventory platform, I've delivered full-stack applications with authentication, CRUD operations, RESTful APIs, SQL databases, role-based access, and cloud deployment. I also trained a LEGO robotics team that earned 2nd Place in the Elementary Division of ROBOCON 2026.