EVERYDAYOS PRINCIPLES

You decide what matters. The system should become more capable without taking that decision away.

EverydayOS is being built around a simple promise: AI should become more useful without making you more dependent on one provider, less in control of your information, or more isolated from the people and relationships your life actually depends on.

Use the best AI for the job

Core product behavior should not depend permanently on one model provider. Better AI should become an upgrade, not a reason to rebuild the system around your life.

Your context belongs to you

The long-term understanding of your projects, people, goals, commitments, decisions, relationships, and history should belong to your operating system rather than to whichever AI model answered last.

Multi-person by design

Real life happens in relationships. EverydayOS should support people and their AI working together through households, businesses, teams, projects, and client relationships while preserving individual context, permissions, authority, and decision rights.

Work and life are both first-class

Projects and clients matter. So do family, health, routines, travel, household responsibilities, and personal goals.

Agents are first-class actors

Specialized AI roles should be able to hold responsibilities, work inside projects, use tools, and delegate under explicit permissions, while remaining downstream of human direction and policy.

Hybrid teams

Humans and AI agents should be able to occupy roles on the same projects and workflows instead of living in parallel systems.

Native projects early

Projects are a core unit of intent, resources, responsibility, and history. They should not live forever inside someone else's project tool.

Progressive absorption

Integrate the tools people already use now, then replace pieces natively only where doing so creates real value.

Stewardship

The system should help steward money, time, attention, human capacity, and compute as scarce resources in service of the priorities the user has chosen.

Cloud first, with a path to more control

The initial product can run in the cloud while preserving a path toward running some or all of the system on infrastructure you control later.

Share only what the relationship needs

Information should keep its source, account, and ownership attached. When context crosses between people or scopes, the default should be the minimum useful disclosure rather than unrestricted sharing.

Mobile early

Capture, voice, notifications, approvals, and everyday-life interaction need to work where life actually happens, not only at a desk.

THE STANDARD

Human agency stays upstream as capability grows.

The more EverydayOS can notice, recommend, prepare, delegate, and eventually do, the more important it becomes that people can understand the context, permissions, boundaries, and tradeoffs behind those actions. More autonomy should reduce coordination load without quietly redefining what matters.

Want to help us build it this way?

Join early access and help test whether these principles hold up against real work, real households, real students, real teams, and the messy decisions of everyday life.

Join Early Access