Structural control for AI work

Let AI help.Keep control.

Dutch Door is a control system for founder-led teams using AI agents. The first door checks whether the work should begin. The second proves it was built right before anything reaches customers or the real world. Between them, AI can move fast without getting the keys to your business.

Born inside LienSuite after the system stopped an unapproved trial-reminder feature before customers saw it.

Pilot work order · customer reminder

A real rule, enforced twice.

Approved work order

One trial reminder, only after a person approves it

The AI receives the goal, limits, and only the safe evidence it needs.

First door

Is this what the founder actually wants?

Conflicting or unclear work is held before the AI begins.

Second door

Is this exactly what was approved?

A reviewer and mechanical checks inspect the result before release.

The AI proposes. A narrow trusted mechanism acts.

The idea

Let intelligence flow. Keep dangerous freedom contained.

A Dutch door makes the boundary obvious: collaboration can pass across the open top while uncontrolled action stays behind the locked bottom.

A split Dutch door with a goat looking over the closed lower half while a sheep, calf, and chicken gather outside.

The familiar metaphor

Open on top. Structurally locked below.

The AI can exchange ideas and return useful work without receiving broad keys to the business. Safety comes from the workshop around the model, not from hoping the model remembers.

Top opens

An approved work order and safe evidence go in; a structured proposal comes back.

Bottom locks

No direct reach into customers, money, private data, credentials, or production.

01

Admit the right work

The first door compares every request with the founder’s rules, goals, allowed actions, and forbidden outcomes.

02

Build in the workshop

The AI prepares a proposal with limited evidence. Approved work is built, reviewed, and tested in isolation.

03

Release one exact action

The second door holds mismatches. A narrow trusted mechanism performs only the approved action and returns a receipt.

Structural protection

Held at the door

  • Work that conflicts with a permanent business rule
  • Customer-facing changes without required approval
  • Requests for unnecessary data, credentials, or capabilities
  • Outputs that differ from the approved work order

What Dutch Door changes

Turn every hard-earned lesson into durable operating leverage.

GitHub made code changes visible, reviewable, and testable before they merged. Dutch Door applies that proposal-and-review model to AI agents and real-world actions—from code and data to customer communication, infrastructure, and money.

When a mistake is found, Dutch Door changes the workshop—narrowing a permission, closing a passage, adding a protected constraint, or strengthening independent review—so the same class of mistake becomes harder to repeat.

Permanent business memory

Hard lessons become tests, protected invariants, and narrow permissions—not another reminder in a prompt.

The first door before work begins

Only approved, unambiguous work enters the controlled workshop.

The second door with a receipt

Independent checks hold mismatches; a trusted mechanism performs the exact authorized action and records what happened.

Private early access

Build the right thing. Build the thing right.

We're inviting founders and small teams already using AI agents to build or run real work—and who need reliable control without becoming AI experts.

Used only for Dutch Door launch updates and your invitation.