LifeOS · Living Office
One office. One agent. All six layers in one room.
The substrate behind every team of agents we ship.
You see the worker. The library, the memory, the brain, the doors — all live in the same room.
What’s inside one office
The image above is one agent at work. Look at the six parts:
The brain
The LLM at the top of the head — Claude, GPT, Gemini, Llama, Qwen. You choose. You can swap tomorrow.
The body
The agent harness on the chest — Claude Code, Pi, Codex, Cursor. The brain plus tools becomes a worker.
The manager
Orchestration — decides who works on what, when. One agent is one worker. A team needs a manager.
The library
Shelves of casefiles · post-its · stories · raw evidence. What the office files. Every claim traces back to its source byte.
The memory
Cards at the top of the shelf. What the agent still knows tomorrow morning — its previous work, its decisions, its lessons.
The doors
Connectors — WhatsApp, Outlook, Slack, Teams, voice. The agent reaches the world without spending cognitive power on plumbing.
One office. Many offices on one engine.
One company is one office — like the one in the image. Multiple companies live side by side in the same dropdown — same engine, same governance, same audit trail, different agents, different libraries.
A team of three on the Mac runs side-by-side with a 250-engineer company on the cloud. Same product. Different scale.
Two brand surfaces of the same product
LifeOS
For individuals.
A team of agents for one person — your work, your library, your office. Personal version.
Living AI Office
For enterprises.
Many people, many agents, one shared library — built for the office floor. Enterprise version.
Same engine underneath. Same six layers. Different surface, different scale, different governance default.
Where it runs
Ordinary infrastructure. Mac and Windows desktop apps · web access in any browser · self-hosted on Azure, AWS, GCP, or your own data centre. Filestore for the library. Postgres for the index. APIs to the brains.