Building structured simulation for geopolitical analysis
Founded in 2024 in Los Angeles. Operating at the intersection of intelligence analysis methodology and large language model capability.
Our Purpose
Decision-makers operating under geopolitical uncertainty face a structural problem: the scenarios most relevant to their decisions are often the ones least well-represented in conventional analysis. Standard scenario planning tends toward modal outcomes and anchors on the recent past — producing analytical outputs that are internally coherent but systematically incomplete.
Principle was founded on a specific premise: that structured computational methods can expand the analytical scenario space available to decision-makers without replacing the human judgment required to act on it. The relevant analytical tradition is not generative AI as productivity tool — it is intelligence analysis methodology, specifically the Structured Analytic Techniques developed within the intelligence community to counter cognitive bias and manage deep uncertainty.
The organization was established in Los Angeles in 2024 to work on this problem at the intersection of intelligence analysis practice and large language model capability. The approach is deliberately narrow: scenario simulation for geopolitical analysis, using structured methods that produce citable, defensible outputs appropriate for institutional decision-making contexts. The aspiration is not to automate geopolitical analysis — it is to give human analysts a structurally more complete view of the scenario space they are working in.
Organization
The Team
Artur Kiulian
Founder & CEO
Artur founded Principle AI to apply structured analytic methodology to geopolitical scenario generation. His work focuses on the epistemics of uncertainty quantification and the integration of computational methods into institutional analytical workflows.
M. Okafor
Senior Scenario Analyst
Specializes in structured analytic technique application and scenario calibration. Background in policy analysis and regional security studies, with a focus on SAT methodology for complex multi-actor environments.
S. Lindqvist
Research Lead
Leads the research programme, including methodology development and publication. Works on probability calibration methods and the application of forecasting literature to geopolitical simulation contexts.
Our Approach
Epistemic rigor
Scenario outputs are only analytically useful if their epistemic status is clear. Principle communicates not just scenarios but the assumptions underlying them, the calibration basis for probability assignments, and the acknowledged limitations of the methodology. An output that looks confident but isn't is worse than a clearly uncertain one.
Calibrated uncertainty
Probability assignments are not decorative — they represent an epistemic claim about the relative plausibility of scenario outcomes. The objective is not to achieve false precision but to produce estimates that are better calibrated than unstructured expert judgment over the ensemble of scenarios considered, using structured methods known to improve calibration in comparable domains.
Structured outputs
Outputs from Principle are designed to be citable, distributable, and subject to analyst review. This means structured formats, explicit assumption inventories, and methodology disclosures — not freeform AI narrative. The audience is analysts and decision-makers who need to account for their analytical process, not end-users seeking generative content.
Analyst augmentation, not replacement
Principle expands the scenario space available to analysts — it does not replace the judgment required to act on that space. The platform's role is to surface scenarios and challenge assumptions that an analyst team might not have considered; the analytical conclusion remains the responsibility of the human analysts using the outputs.
Contact us
For access requests, partnership discussions, or research inquiries, reach us through the contact form.
Contact Us