From Physical AI demos to real-world deployment
My conversation with Edison Smart for The Edison Edge on what it takes to move Physical AI beyond a compelling demo and into reliable real-world operation.
Notes
Products, architecture decisions, open-source contributions, and lessons from shipping.
Archive
7 posts covering software products, AI, open source, and engineering systems.
My conversation with Edison Smart for The Edison Edge on what it takes to move Physical AI beyond a compelling demo and into reliable real-world operation.
How separate personal and work knowledge systems make AI agents more useful while keeping context, permissions, and confidential information safely bounded.
How a collision mesh workflow I needed became a merged contribution to URDF Kitchen, and why contributing improvements back matters.
A tiny macOS menu-bar app that came out of a week of agentic robotics workflows and turns ordinary keystrokes into piano notes.
At the Google DeepMind Hackathon Chicago 2026, I built Amdion: a local, context-aware desktop assistant that can understand screens, transcribe voice commands, and execute actions across apps.
How GitHub Copilot helped improve this blog by converting posts to markdown and streamlining the development process.
This website/blog comes from my continuous interest in learning new technologies and sharing my experience in the hopes of it being beneficial for any reader.