From complexity to action

We develop and apply a range of methods - AI, machine learning, mechanistic models - to represent systems. We focus on choosing what's appropriate, not what's fashionable.

We design for iterative validation and improvement, choosing appropriate metrics and placing emphasis on stakeholder co-creation.

We characterise different forms of uncertainty: what is genuine variability, what is missing knowledge, and where biased perspectives have created blind spots. 

Often these forms are collapsed into a single range or worse, a single estimate.

Separating them shows where uncertainty can be reduced and whether stated confidence holds up against what actually happens.

We synthesise diverse forms of evidence to prepare for uncertain futures. 

Rather than giving static advice on a one-off choice, we provide insight and resources for our clients to optimise decisions on an ongoing basis, as the context shifts.