The engines · RDA

Risk Diagnostics maps the risk spectrum.

What could happen? · ranges, not a commitment

Your team can hand-build a few scenarios before a meeting — best, base, worst — then guess at everything in between. Risk Diagnostics runs hundreds of plausible futures, solves each, and shows you exactly where you’re exposed.

The deepest assumption

Drivers are shapes, not numbers.

The average hides the expensive days. A driver’s shape — bimodal, fat-tailed or tight — is what a resilient decision is actually defended against. Each is learned as a distribution from your own transaction history.

BimodalMean ≠ typical

Two regimes — a fast lane and a slow lane. The mean lands in the empty valley that rarely happens.

Fat-tailedTail drags the mean

Most days cluster early, but a long right tail of expensive days drags the mean past the peak — the average overstates the typical week.

TightNumber is honest

Every future lands within a hair of the mean. Here the single number is honest — but you only know that once you’ve seen the shape.

The seeing layer

Variability enters the engine in its true shape.

Before anything is decided, the world is read honestly. This is what changes in the first weeks of a deployment — and what every downstream range depends on.

Distributions, not values

Lead time, demand, yield and cost enter as the shapes they actually have — learned per item, supplier, location and customer — not as a single static number that is wrong a different way every week.

Order lines, not buckets

Demand and supply enter at the grain decisions are made on — order lines, histories, real partners — not aggregated into buckets that lose whatever intelligence the detail held.

Signals, not surveys

Short-horizon sensing, causal structure and ecosystem signals enter as first-class inputs, so the engine reads the forces reshaping demand now rather than extrapolating the past.

At a scale no team can do by hand

Every plausible future, mapped.

Risk Diagnostics learns the shape of every driver and risk, samples them into many futures, and solves each — one optimal plan per future, sweeping the risk probabilities. The result is the full spectrum of KPI outcomes.

  • ·Calibrates variability from your own history
  • ·Runs many scenarios — one optimal solve per future
  • ·Produces KPI ranges, bands and resilience metrics
  • ·Reveals which drivers move which KPIs
The diagnosis

Risk Diagnostics doesn’t tell you what to do. It tells you everything that could happen — so nothing surprises you. The decision comes next, in Resilience Optimization.

From raw history to rich futures

Driver calibration, in four steps.

Take any numerical object in the model, learn its shape from your own history, then choose how it is applied to build every future.

  1. 01

    Any numerical object

    Supplier lead time, demand, scrap rate, capacity, energy cost — anything the model carries.

  2. 02

    Learn from past history

    Observed in the transaction history of your source systems — ERP, CRM, MES.

  3. 03

    Driver variability histogram

    Its real distribution, per item, supplier, location and customer.

  4. 04

    Stitch into iterations

    Choose how each driver is applied — replace, add, multiply — to generate rich, realistic futures.