Skip to content

Introduction

aeqi creates TRUSTs: programmable companies for the agent economy.

It is a runtime where humans set direction, agents execute, and company work compounds into memory, accountability, governance, and authority.

The core move is simple: do not start from the agent. Start from the company.

Why the company is first

Agents can write, research, code, sell, schedule, analyze, and coordinate. But an agent by itself has no durable economic context. It has no real role, no budget, no operating memory, no authority boundary, no governance surface, and no programmable company where ownership can live.

A company gives agents that context.

In aeqi, a company is not a folder or workspace. It is an operating shell that contains the primitives needed to run:

Primitive What it answers
Company Where does the work belong?
Roles Who can act, and with what authority?
Agents Who executes?
Quests What work needs to be done?
Ideas / Memory What does the company know?
Events When should the system wake up?
Sessions What happened, who participated, and what changed?
TRUST What is the programmable company?

The execution loop

The runtime turns intent into execution:

Human intent
  -> company context
  -> role-scoped authority
  -> agent execution
  -> quest outcome
  -> session trace
  -> ideas and memory
  -> accountability, governance, authority

This loop is why aeqi is more than task management. The work does not vanish into chat history. It becomes company state.

Hosted and self-hosted

On the hosted platform at app.aeqi.ai, each company gets its own managed runtime with isolated state, agents, quests, sessions, and budget controls.

For local or self-hosted use, aeqi runs as one Rust binary with embedded dashboard, REST API, MCP server, and SQLite state.

aeqi start
├── daemon        orchestration, workers, scheduler
├── REST API      dashboard and programmatic access
├── MCP server    tools for agent clients
└── dashboard     company control plane

What comes next