One organization. Two divisions. A single argument about intelligence.

The industry is full of people who can call an AI API and few who can explain what happens when they do. CAS exists to close that gap in both directions: CAS Solutions builds advanced intelligent systems; CAS Studies develops the engineers capable of designing them.

They are not two businesses. They are one ecosystem where research, production and education reinforce one another. What we deploy on weekdays edits what we teach; what we teach constrains what we are able to deploy.

We treat intelligence as a stack — mathematics, representations, models, perception and language, foundation models, generative systems, agents, world models — where each layer is explicable and engineerable. Mysticism is a failure of explanation, and we do not accept it.

That view is why ModLens exists: architectures should be seeable. It is why our curriculum begins with linear algebra and ends with autonomous systems. It is why our proposals state assumptions and our studies report negative results.