HomeMah
Building an AI-assisted interior design application from the renderer upward.
01 / Context
A custom desktop 3D interior design application developed with C++, Qt and OpenGL, combining scene editing, asset workflows, persistence and local AI/RAG experiments.
02 / The problem
Interior design tools are a useful way to explore several difficult systems at once: rendering, interaction, scene state, asset loading and user-facing editing. I wanted to understand those layers directly rather than hiding them behind a game engine.
03 / What I did
HomeMah was built around a custom desktop editor. I worked on the rendering and scene pipeline, furniture and material handling, editor controls, JSON-based persistence, packaging, and experiments with locally hosted AI through Ollama and RAG infrastructure.
04 / Where it landed
The project became a working Windows application and a continuing testbed for graphics, tooling and AI integration. The portfolio case study focuses on what is implemented today and keeps future AI ideas clearly separated from shipped functionality.