Skip to main content
friday-ai.dev

Many ideas.
One Friday.

I build software around the things I care about, from off-road communities to AI tools. Project Friday is the environment I’m building to connect that work: memory, workspaces and AI, shaped around how I think and create.

Illustrative creator’s workspace connecting off-road community software, photography and coding through Friday.

Personal projects. Shared knowledge. Tools I build, and tools I build with.

Friday-ai.
Keep what you learn.

A new development session should not mean starting from scratch. Friday-ai brings project lessons, decisions and working commands into a short brief, so the next agent has useful context before it starts.

I can explore the connections between ideas, review what the system remembers and keep the knowledge that earns its place.

  • Session briefs
  • Project lessons
  • Reviewed memory
  • Knowledge graph
Conceptual illustration of connected knowledge above layers of project records.

Sandcastle.
A place to build.

An idea needs somewhere to take shape. Sandcastle gives an AI agent a disposable development workspace to build, test and preview software. Each task gets its own environment, with cleanup when the work is done.

Live previews let me see the work as it develops. Temporary CI runners support the checks that help turn an experiment into a working project.

  • Development workspaces
  • Live previews
  • Automatic cleanup
  • CI runners
Conceptual illustration of separate, temporary development environments.

§ Voice

What Friday-nexus will be.

Mid-build · v2.2

Honest section. What’s here is in motion.

Voice is the interface I want, mid-build. Wake-word activated and running. Cross-platform via Tauri 2: desktop on macOS, Windows, and Linux, live-validated on Windows so far. Mobile (iOS, Android) is planned, not started. Speech streams through Pipecat, which routes each turn down one of two paths: a fast path on Gemini for quick replies, and a slow path that hands the turn to a Claude agent able to reach the brain and the session runners over MCP. Either way the answer comes back as speech.

The voice loop runs end to end for real now, not just in fixtures: I say the wake word, Friday answers out loud, and I can talk over her to cut a reply short. Barge-in took a few rounds to get right. The avatar caught up too. What renders today is a reactive orb that pulses with the voice, built in WebGL2. A textured 3D head, a VRM model, is built behind it and waiting on visual validation before it goes on screen. The 3D avatar the name implies is closer than it was, still not shipped.

The name comes from a different fictional AI. I’m not aspiring to Iron Man, just to a stack I can talk to without typing.

Friday-nexus: a glowing translucent voice orb with a faint face forming inside, ringed by concentric audio waveforms.

What’s inside

  • Wake-word: openWakeWord, always-on, runs local
  • Cross-platform: Tauri 2 desktop (macOS, Windows, Linux), live on Windows; mobile planned, not started
  • Orchestrator: Pipecat, streams speech in and out
  • Fast path: Gemini, for quick replies
  • Slow path: a Claude agent reaching the brain and sessions over MCP
  • Avatar (today): reactive WebGL2 orb, voice-driven, live today
  • Avatar (next): textured 3D head (VRM), built, visual validation pending

Friday-nexus

Local AI.
Connected to my projects.

I’m building a local AI environment designed to bring coding assistance, reasoning, project-memory review, image understanding and video creation into one connected workflow, with selected AI workloads running on infrastructure I control.

Illustrative local AI workstation with conversation, image understanding and video tools.

Build and reason

A place for coding assistants and general models in the development process, from exploring a problem to examining an approach.

Understand and organise

A role for visual understanding, memory review and specialist classification, helping turn information into useful context.

Create with AI

A creative direction that connects ideas, visual inputs and local video generation within the same personal environment.

Different projects.
A common thread.

I build around real interests and problems: bringing an off-road community together, exploring research, organising knowledge and finding new ways to work with AI.

Friday connects the way I work on them. Useful project knowledge gives the next task context. A workspace gives the idea room to develop. Review helps decide what to keep and what to improve.

Some projects connect directly to Friday. Others share the development practices and tools. Together, they reflect the way I turn an idea into something useful.

  1. Start with a question

    Bring an idea and the project knowledge that helps make sense of it.

  2. Build something useful

    Work with AI agents, preview the result and test it against the problem.

  3. Carry the lesson forward

    Review the work and keep useful knowledge for the next session.

See the projects

Ideas, made tangible.

A selection of my projects and collaborations. Different purposes, connected by the curiosity that gets me building.

Explore all projects