Noesis is the OS for decision sandboxes. Describe an environment, and a population of agents comes alive inside it — so a person or a company can rehearse any decision, and explore the whole space of moves, before committing anything real.
Each sandbox fixes a scenario and swaps in its own population, actions and outcomes — the same multi-agent kernel runs underneath. Open one and it runs, right now, in your browser.
Not a chart that predicts one number. A living world you can poke, re-run and search — where the outcome emerges from thousands of agents actually behaving.
A person or company designs the environment they want to understand. What actually runs inside it is a society of agents. The two form one closed loop — that is the whole product.
Set the population, the actions on the table, the outcomes you care about. In plain language, or with the levers. This is the human-facing side.
Thousands of heterogeneous agents perceive, decide and act — with memory, taste and fatigue. The outcome you read off is what they did. Agents can drive it too, via the same kernel.
One coherent loop, from a question to a move you can defend — and you can run it as many times as you like.
Open the live sandbox — it runs in your browser, right now.