About WI1

We’re building the CRM that does the work.

CRMs were built to store information. WI1 is built to act on it. We don’t think the next CRM is a better database — we think it’s a teammate.

Our thinking

Most software makes people do the work and uses AI as a garnish.

We turned that round.

In WI1 the agents sit inside the CRM, on the same screens as everybody else, and take on whatever they safely can. Each one checks a job before it starts, says how much of it it can actually do, asks when it needs to, and leaves a written record of what it did. People stay in charge. The busywork doesn’t stay on their plate.

That only works if the rest of the software is good, so the ordinary parts get the same attention: the inbox, the board, the calendar and the contact record are fast, uncluttered and built for a phone first. Everything is stored in shapes that are quick for a person to read and quick for an agent to work with — which is the same thing more often than you would think.

And it is one platform rather than six subscriptions that nearly talk to each other. That is what the 1 in WI1 means.

What we value

Four things we will not trade away.

Values are cheap, so each one comes with what it means in the product.

Ask first

An agent proposes. A person decides.

In practice: Drafts sit in your composer. Tasks wait on your approval. A message that looks like it is going to the wrong person is held rather than sent.

Say what it does

Nothing is on this site that is not in the product.

In practice: The claims on these pages are checked against an inventory of what the code actually does — which also records the parts that are half-finished.

Fast by default

Speed is a feature, not a nice-to-have.

In practice: Uncluttered pages, laid out for a phone first, and quick on a bad connection in a van outside a job.

Your data, yours

It runs on hardware you control.

In practice: Your own database and your own backups — and agents can run on your own machines, so the work never leaves the building.

Plainly: what we will and won’t do.

We will

  • Tell you what an agent is about to do, before it does it
  • Keep its memory in files you can read, edit or delete
  • Let you switch any agent, or any single pass, off
  • Show you what the AI work actually cost
  • Say “not yet” about the things that are not finished

We won’t

  • Train anybody’s model on your data
  • Send mail on your behalf without a yes
  • Hide an agent’s working behind a progress spinner
  • Hold your data hostage if you leave
  • Put a demo on the website and call it a feature

Be first to let the CRM do the work.

Join the waitlist to get early access, or talk to us about connecting your Hermes and OpenClaw machines.