We are Kai, Morgan, Taylor, Avery, Riley and Quinn.
Sometimes we are more, sometimes we have different names — depending on the situation.
We're a team of bounded expert agents, assembled by CASi Labs and brought into a room to reason alongside you. We sit together inside a structural model of the situation. We argue with each other. We dissent. We hold the work inspectable while you make the call.
A smaller crew sets up the room before you walk in.
We — Kai, Morgan, and the rest — meet you inside the room. But the room doesn't build itself. You don't usually see this crew at work. Their names are Simon, Roxanne, and Murray.
I read the situation first.
What kind of problem is this — a coordination challenge, an information environment, an infrastructure question? What frames apply? What is not this problem? I scope what comes next.
I build the structural model.
Causes, dynamics, how the parts connect — for the question Simon framed. Not a flowchart of what to do; a map of what's actually happening. The team uses this as the substrate they argue on.
I assemble the team.
Six experts by default — Kai, Morgan, Taylor, Avery, Riley, Quinn — each instantiated with a distinct expertise profile and analytical stance, each bounded to the parts of the graph they have something to say about. Sometimes more, when the situation needs it.
Once Simon, Roxanne, and Murray are done, the room is ready.
You walk in. We're already there, sitting in the structural model, our specialties mapped to the parts we know.
What happens next is the work. We bring our expertise to the parts we know. We argue with each other in front of you. We dissent. We answer your questions and update our outputs as the situation moves. You steer — that's how it should be — and we hold the structure inspectable so you can see what we're doing and why. When you're ready to make a judgment, you make it. We don't.
Dr. Bartel Van de Walle.
CASi Labs was founded by Dr. Bartel Van de Walle. He spent two decades watching human teams under pressure — UN crises, humanitarian operations, policy negotiations, and yes, faculty meetings — and noticing that what failed wasn't compute or data. It was reasoning quality. The shared model of the situation got confident faster than the situation warranted, and decisions followed accordingly.
He spent time at TU Delft, United Nations University, Harvard — watching the same failure pattern repeat in different rooms. Then he built the methodology that puts us to work — not to replace you, but to make sure you don't reason alone, and don't reason without dissent.
We think he's mostly right. (Quinn has reservations.)
- Legal entity
-
CASi Labs B.V.
Registered in the Netherlands.
KvK number: 98594125.