I designed a workflow system that helps engineers move from scattered product context to structured software delivery.

The system connects requirements, design references, repository context, implementation planning, validation, audit documentation, and pull request delivery into a reviewable engineering flow.

The goal is not to replace engineers. The goal is to reduce repetitive coordination work while keeping human judgment, architecture review, and accountability at the center of the process.

Large engineering systems are hard to understand because knowledge is scattered across repositories, documentation, logs, and team memory.

I built a repository intelligence workflow that indexes codebase context, generates searchable engineering knowledge, and helps connect product feedback to potential engineering actions.

One use case is free-text user feedback classification — identifying whether customer feedback may indicate a bug, enhancement request, or non-defect signal.

mAI is a long-term exploration of what an AI teammate for musicians could become.

The idea is to help musicians compose with less friction: capturing MIDI input, translating performance into structured musical representation, analyzing melody and harmony, and eventually suggesting musical ideas in real time.

This project combines my background as a software engineer, musician, and composer.