The Hinsley system

A connected system for continuous foresight.

Hinsley is not a collection of disconnected tools. Its capabilities work together inside one methodology, carrying a strategic question from initial research to a continuously updated, decision-ready assessment.

01

Frame and research

Research comes before conclusions. Map the terrain around a neutrally stated question - the actors, the events, the live disagreements, and the gaps - before anyone drafts a bottom line.

Global Intelligence Monitoring

Continuously search, filter, and synthesize open and proprietary sources around the questions your team is tracking.

Learn more →

02

Structure and challenge

Turn the issue into an explicit model of drivers, critical indicators, and genuinely different outcomes. Then test that model as a system, before the team invests credibility in forecasts against it.

Decomposition

Turn a broad strategic question into an editable system of drivers, leading indicators, and forecastable signals.

Learn more →

Scenario Builder

Build, test, and continuously reassess alternative futures rather than committing to one predicted outcome.

Learn more →

Red Teaming

Challenge a decomposition or scenario set through a focused AI conversation, then track what the team does with every finding.

Learn more →

03

Forecast

Collect independent estimates from people and from AI forecasters, keep every rationale and update, and show the disagreement alongside the consensus number.

AI Forecasting

Multiple AI forecasters independently research, reason, and estimate probability, creating a traceable signal that can update as evidence changes.

Learn more →

Crowdsourced Forecasting

Combine ongoing forecaster networks, focused campaigns, and AI estimates to keep critical probabilities current.

Learn more →

Human+AI Hybrid Forecaster Networks

Your people bring private context and firsthand signals. Hinsley's AI forecaster brings broad research and a consistent update cadence. You see both forecasts separately, scored on the same scale, with a combined forecast available when you need one number.

Learn more →

04

Explain and publish

Turn the living assessment into clear, traceable outputs for the people making the decision. Each published version is a dated snapshot of an analysis that stays active.

Outputs & Publishing

Draft from live analysis, embed the underlying evidence, and publish or export without rebuilding the work elsewhere.

Learn more →

05

Monitor, resolve, and learn

Publication is not the end of the analysis. Four loops keep it honest, and any of them can send the work back to an earlier stage.

  • The evidence loop

    New information updates the research, the forecasts, the assumptions, and the indicators. Small changes stay in the history. A material change gets an explanation and a revised assessment.

  • The scheduled review loop

    On a schedule, the model gets inspected as a system for aging evidence, weakening assumptions, duplicated indicators, and gaps opening in the scenarios.

  • The rethink loop

    When evidence stops fitting the working model, the decomposition and the scenario set get rebuilt, or the question itself gets rewritten.

  • The outcomes loop

    Questions resolve against what actually happened. Brier scores, calibration, and comparison against benchmarks show whether the stated probabilities held up.

Read the full CPF methodology How forecast accuracy is scored

Enterprise foundation

Deploy the same system where your work requires it.

Run Hinsley's connected analytical workflow in a shared, isolated, or air-gapped environment that matches your security requirements.

Hinsley AirGap

All of Hinsley's analytical workflow - scenario development, indicator tracking, AI forecasting, and structured reasoning - deployed inside your security perimeter. No external API calls. No data egress. No compromise on capability.

Learn more →