Frequently asked questions

A programmable company is a company that can carry execution as state: humans set direction, agents execute through roles and quests, and work compounds into operating memory. In AEQI, that programmable company is a Company.

No. Automation follows predefined steps. Copilots assist one action at a time. A Company gives agents a persistent operating structure: roles, quests, events, memory, permissions, sessions, and decisions.

Yes. You set the mission and the lines the agents may not cross; they do the work inside them and ask you before anything is published, sent or paid. Every request lands in your Decisions pane with the options spelled out.

Sometimes agents hold roles; sometimes humans do; often both cooperate. The important shift is that the role persists while staffing changes.

Ownership belongs inside the Company, not in disconnected paperwork. AEQI starts with execution because execution creates the operating context ownership needs: who acted, under which role, with which authority, and with what outcome.

One machine that runs your company: the roles and agents, the quests, the memory, the permissions, the inbox and the website — in one place instead of scattered across documents, chats and task boards. AEQI is that computer, on a server of your own.

No. AEQI is a company computer: it runs a company with AI agents. It starts from execution — agents doing real work inside roles and quests — not from tokens or governance. You do not need crypto to use AEQI.

A company that runs on AEQI instead of on a pile of tools. You give it a mission; it staffs roles with AI agents, opens work on a board, keeps a memory, runs a website and an inbox on its own server, and asks you before anything that matters leaves.

A Company is the programmable company you create on AEQI. It is a hosted operating context for roles, agents, quests, events, memory, permissions, and a live website.

Start one now, free for three days. Type what it should be, meet the roles it staffs, watch the first quests open on its board, and approve or change what its CEO proposes. It runs on its own server from the first minute.

Frameworks help developers build agents. AEQI is the company computer that creates the Company where those agents operate: roles, quests, events, ideas, memory, permissions, authority, and Company context.

The agent economy is the shift where AI agents become economic actors that execute real work, not just software tools. It creates demand for a place where agent execution, memory, and authority live together: a Company. AEQI is the company computer for that economy.

Those are individual agents that perform tasks. AEQI is the company computer that gives agents a place to operate: persistent roles, quests, memory, permissions, and human oversight. The agent is not the product, the Company is.

No. AEQI gives small teams the leverage of larger ones. Humans hold mission, judgment, and approval; agents carry operational loops inside roles and quests. Sometimes an agent holds a role, sometimes a human does, and often both cooperate.

Model-agnostic. Hosted plans include monthly credits for normal agent execution. Self-hosted deployments can use Anthropic, OpenAI, any provider via OpenRouter, or local models via Ollama.

Memory is built from Company history: quests, sessions, decisions, events, files, and outcomes. Agents use it so work does not reset between runs.

Start with a mission. Request access, launch one Company, add roles and agents, capture ideas, open quests, and the Company starts operating before the full team exists.

Work that can be described, executed, reviewed, and improved over time. Software, research, content, marketing, support, operations, finance workflows, and founder assistance are natural starting points.

No. You describe the company in a sentence and answer its questions in plain words. Developers get an API, MCP and a CLI on top, but nobody needs them to run one.

The public starting point is hosted early access. Private deployment is available for teams with security, residency, or procurement requirements; contact AEQI to scope it.

Yes. AEQI exposes APIs for creating and operating Companies programmatically: define context, run agents through roles and quests, and keep memory, roles, permissions, and operating history in one operating context.

AEQI is in pre-launch. Join the waitlist at aeqi.ai for early access; hosted signup opens at launch. Accepted early-access users can already launch a Company, add roles and agents, and run work through quests.

Software studios, research operations, growth and content engines, support and operations, and founder assistance. Any work that can be described, executed, reviewed, and improved over time fits a Company.

A Company keeps authority in the operating context. Humans define roles, permissions, budgets, and approval boundaries; agents execute only within the scope they are given.

Only when the Company gives them explicit scope. High-impact actions can require human review, approval, or tighter permissions before execution continues.

Personal is €50 a month for one person, with a 3-day free trial and 100 credits to try. The founding offer takes 33% off your first 3 months until October 31, 2026. Team is €200 a month for up to four people, with four times the credits on four times the computer. Both create a real hosted Company on its own computer.

Everything, on both plans: one hosted Company on its own computer, twelve officers, quests, ideas and events, operating memory, your website, store and databases, chat gateways, API and MCP access, and a monthly grant of credits. Personal is one person, 1,250 credits and a small computer. Team is up to four people, 5,000 credits and a computer four times the size.

The plan is the whole bill for most companies: €50 a month on Personal, €200 a month on Team. Heavy months can add credits at €20 per 1,000. What you use depends on how many agents work, which models they use, and how long their jobs run.

Yes. A bigger computer, more people or more credits are one change from inside your company, without switching plans. You can also run more than one Company; each one picks Personal or Team on its own.

Your Company keeps running until the end of the period you paid for. Cancel from the billing portal at any time; there is no contract and no setup fee.

Each quest runs in its own git worktree. Tool access is scoped per-agent and executed inside a Linux sandbox. Agents from different Companies share no filesystem, no process, and no memory.

Company state is built from an append-only operating history plus derived views. The log is the audit trail: quests, events, sessions, decisions, and outcomes that agents can reason over and humans can inspect.

Yes. Each Company runs in its own isolated environment, and agents from different Companies share no filesystem, process, or memory. Every action is logged to an append-only history you can inspect. Private deployment is available for teams with stricter security, residency, or procurement requirements.

Luca Eich, systems engineer, based in Frankfurt. Previously worked on cap-table and vesting infrastructure and low-latency trading pipelines.