Governed Decision Intelligence

The seat no one
else occupies.

ARACHNE is not a planner. It is the governance layer that sits above the planners and the models — the place that decides whether the decision can be trusted. It doesn't replace your planning system. It multiplies it. We named the category because the category didn't exist.

Where it sits

On top. Not instead of.

The floor
Planning systems

Advanced Planning Systems concurrently re-plan supply at scale. Best-in-class at what they do. ARACHNE consumes their scenarios; it does not compete with them.

The noise
Black-box AI

Models that assert an answer with a confidence score that rarely means abstention, and an audit trail that is often a screenshot and a hope.

The governance layer
ARACHNE v30.5

Takes what the floor produces and governs it — proving the record, naming the gaps, refusing to clear a decision that hasn't earned it. The seat above the stack.

What the governance layer adds

Above the category line.

Not "our planner beats their planner." A different job entirely — the properties a governance layer must have, which the layers below it were never designed to provide.

Property of a trustworthy decision ARACHNE v30.5 Planning-system category Black-box AI
An explicit "I don't know — here's what I'd need" state ● First-class Returns a plan, not an abstention Asserts confidence; rarely abstains
A safety layer that can only block, never authorize ● By construction Not an architectural guarantee
A tamper-evident, sealed record of every decision ● Hash-chained Audit logs, not sealed chains Trail is a screenshot and a hope
Refuses to clear under unresolved contradiction ● Escalates Resolves to a recommendation
The explanation is the computation — not narrated after the fact ● Reproducible & gated Explainability bolted on after
The minimal, cost-ranked evidence that would change the verdict ● PRESCIENT

The middle two columns describe tool categories — honest generalizations, not audited measurements of any specific product. Kinaxis® and Blue Yonder® are trademarks of their respective owners; ARACHNE is independent and is not affiliated with, endorsed by, or certified by them.

The differential

A moat made of math,
not features.

  • PRESCIENT. The minimal, exact, cost-ranked evidence set that would move a conclusion to a target verdict — in four modes, including "structurally unreachable" and "ignorance, here's the data I'd need." It is not a feature you bolt onto an optimizer or a language model.
  • Interlocking integrity guarantees. A single copied feature is a screenshot. The differential is a set of guarantees that hold together — a sealed genealogy, a safety layer that can only block, provenance tiers with no path from synthetic to prospective, and a cap on machine-authored conclusions certifying themselves.
  • Building toward post-quantum. The architecture is designed toward post-quantum capability and adversarial stress resistance — a differential in the thinking, not a claim of arrival. Most of the layer below isn't architected for it at all.
  • Timing. The day provability becomes a buying requirement, the layer that can prove is the one already standing on a hardened substrate.
Read the fine print — we do

Where it's measured,
it's scoped.

Complementary, not competitive with process frameworks. SCOR describes what the chain is; ARACHNE governs what to decide when the evidence conflicts. They are layers, not rivals.

One measured figure, its exact scope. In a pharmaceutical signal-detection benchmark (a 538-case corpus), the mechanism scored F1 0.942 against published academic methods. That is a mechanism measurement in one domain — not a supply-chain claim, and not field efficacy. Field efficacy remains unclaimed, by design.

Mechanism · measured Field efficacy · ignorance
Provenance — metric: F1 0.942 · corpus: 538-case pharmaceutical signal-detection · comparison: published academic methods · scope: pharma only, not supply chain. Efficacy resolves with measured field deployments.