Public lab By Mads Bach Andersen

I follow questions
until they become tangible.

Some questions become companies, products or prototypes. Others leave a useful model—or a clear reason to stop.

Current question · Systems that act

When software acts,
reality needs a record.

When software acts for us, every step should be inspectable. People need to know what happened, why—and how to reverse it.

Drawn from Trase · SYNK · Loopable · Connected Wear

Recurring areas

Three areas
I keep returning to.

Each connects earlier work with questions I am still exploring.

Systems that act

When software takes action for us, how do we stay informed, in control and able to take it back?

Systems that act. Bounded action leaves a trail someone can inspect. A request passes through a cobalt permission gate, becomes an action, prints an inspectable receipt, and returns along a green undo path.
Bounded action leaves a trail someone can inspect.

Related work and research

  • Traseproduct · stopped
  • SYNKresearch · dormant
  • Loopableresearch · archived
  • Rental + payment agentsresearch · archived

Technology close to us

As technology moves onto the body and into the room, how can it become more useful while asking for less attention?

Technology close to us. The interface moves from screen to body to place. Dense signals leave a screen, soften as they cross a human body, and dissolve into one quiet cue in the surrounding room.
The interface moves from screen to body to place.

Related work and research

  • AR + spatial computingresearch · recurring
  • PodVoiceprototype · built
  • Connected Wearresearch · archived
  • Huddling Peakconcept · rejected

Everyday infrastructure

Where can housing, local services and everyday exchange lose friction without losing trust?

Everyday infrastructure. Value moves through everyday systems—and sometimes comes back. A trusted route connects a home, local service, exchange point and provider; repeated proof tokens continue around an obstruction by a recoverable return path.
Value moves through everyday systems—and sometimes comes back.

Related work and research

  • Keyholecompany · operated
  • Gloovecompany · operated
  • Urbanizedconcept · archived
  • Rent2Cashconcept · unbuilt
  • Urban storagestudy · not pursued

Past work

What became real

Two companies and one product
that moved beyond an idea.

Gloove

A moving-services company Mads founded.

It raised capital, launched and served customers for several years. The unit economics did not hold.

Company · operated

Keyhole

A rental-technology company Mads co-founded.

It raised capital and reached distribution; the category economics stayed difficult.

Company · operated

Trase

An invoice-retrieval product, built and then stopped.

The retrieval feature worked; the problem was too narrow to support a company.

Read the Trase record

Product · stopped

How the lab works

Curiosity becomes
material here.

The point is not constant output. It is to leave each question with something useful: a model, a test or a decision.

  1. 01

    Research

    Look wider than the first answer.

  2. 02

    Model

    Make the system visible enough to challenge.

  3. 03

    Test

    Build the smallest thing that can prove it wrong.

  4. 04

    Decide

    Continue with evidence—or stop cleanly.