Products
What we are building
We build our own AI products, and they are in development now. Each one is an AI agent that does real work in the tools people already use. A person approves what matters, and the data stays where its owner puts it.
The kind of product we build
Real work in the tools you already use
The agent does a whole task inside the email, documents and business systems people use today. Nobody has to move their work to a new screen first.
A person approves what matters
The agent prepares the work, then waits. Anything with consequences, such as sending, paying or changing a record, goes ahead only when a person approves it. Every step is recorded.
Private by design
Data stays where its owner puts it: on their own servers or in the cloud region they choose. The products are designed that way from the start, not changed later.
How we build products
Our products go through the same harness and the same checks as all our work.
How we buildThe same harness
AI coding agents work inside the harness we maintain: written project rules, and a plan, build and review pipeline. A person owns each piece of work and decides what ships.
Checks that can fail
Every change runs through tests, type checks and a build gate. A fresh reviewer then tries to break it. The agent that wrote a change does not judge it.
Shipped when it passes
A product ships when it passes those checks, not when a date arrives. If a check fails, it is not ready yet.
Status
In development. We will share names when they are ready.
Get updates
Leave your email address and we will write when there is something to show.