This is where I log experiments in robotics, embedded systems, AI and AI agents, and personal infrastructure. Public content is open for browsing; privileged tools are only reachable inside my WireGuard mesh.
Embedded systems and AI hobbyist; robot and agent infrastructure tinkerer.
Internal · WireGuard ◫Renders the skills repository into an AI-friendly technical docs site. First piece: Remote Development Environment (a WireGuard mesh).
Public ◎Notable news and product releases in robotics, AI and embedded systems, curated.
Public ✓Kanban for todos, deadlines and priorities; jot things down in natural language.
Internal · WireGuard ✦One entry point for sessions and remote terminals: browse history, start a new session, or connect straight to the machines.
Internal · WireGuard ❖Files assets, settings and finished pieces into separate workspaces — one place to push them forward, one place to keep them.
Internal · WireGuard ⌁Online status of WireGuard nodes and key services.
Internal · WireGuard ◐Describe the picture or the line you want and a high-performance GPU renders or voices it; switch between workflows, characters, voices and styles.
Internal · WireGuard ✧Dataset upload, auto-tagging and live progress; collect a trained model whenever you like and download it in one click.
Internal · WireGuard ▦A continuously evolving experimental space, updated from time to time.
Internal · WireGuard ⇩Toolchains, device support packs, hardware manuals and prebuilt artifacts, grouped by purpose.
Public ✉Sign in to send me a message that only I can read; the public message board isn't open yet.
Public ✱Open-source AI and robotics projects we actually use.
Intranet · WireGuard ¶Good papers I've read, with my breakdown notes, kept as a browsable list.
In the works ◉A spot for interesting visual, data and interaction experiments later on.
More to comeTactile sensing, cameras, inter-processor communication and time synchronization.
Chassis, dual arms, waist, lift and a unified Zenoh bus.
Cross-device session sync, daily work reports, the task center and a desktop pet entry point.
Image, video and speech generation, style and character LoRAs, and a browsing UI that works on a phone.
Dataset upload, auto-tagging, training monitoring and one-click model collection, all done in the browser.
A new direction in the works; still working out the details.