We build systems that observe, think, and act inside real operations.

Our focus is on systems that sit close to the business: they watch what is happening, form opinions, and take actions. Not tools, not dashboards.

Infrastructure for decisions and execution.

Post-agentic systems
Operational intelligence
Work

Where we apply these systems

We work in domains where decisions are frequent, the stakes are real, and systems are allowed to sit at the core of how work gets done.

I
Cognition systems

Architectures for systems that carry state, reason over time, and revise their own conclusions as new information arrives.

II
Operational decision loops

Systems that watch a domain (markets, operations, risk) and suggest or take actions based on evolving policies and constraints.

III
Autonomous marketing

A long-term effort to give marketing a body: perception, strategy, execution, and learning running as one continuous system.

Read the marketing system brief →
Principles

How we approach the work

A few constraints that shape how systems are designed, deployed, and allowed to evolve.

Start from the loop: signal, interpretation, action, feedback.
Keep the visible surface simple; put the complexity in the core.
Introduce autonomy gradually with clear scopes and guardrails.
Prove value in production, not in decks or staged demos.
If this way of working matches a real problem you have, get in touch and we can see if it is worth exploring.
Send an introduction