The layer · SOI / SOP / SOR

You can’t patch this into a planning system.

A new layer — above the stack you already run

A record system holds what happened. A planning system turns intent into a feasible plan. Neither was built to decide across a range of futures. That is a different job — and it needs a layer of its own.

Three systems

The decision layer sits on top.

The layers below stay exactly where they are. What changes is what they’re allowed to produce: not a single brittle plan, but a decision that already accounted for the futures the plan never saw.

SOISystem of Intelligence
Decides

Chooses the defensible action across futures. This is the layer VYAN introduces. It is interoperable by design — connectors exist for every major planning and record system, and it runs standalone when you’d rather it did.

SOPSystem of Planning
Plans

Turns intent into feasible plans and schedules.

SORSystem of Record
Records

Holds transactions and master data.

Why the old layer leaks

A paradigm engineered for when volatility was the exception.

Four assumptions held when the world moved slowly. Each has expired — and a faster planning machine only amplifies the leak, because the leak is structural, not predictive.

01

Sequential S&OP

Demand → Supply → Production → Distribution. Each step inherits a hand-off it can no longer challenge.

02

Silo optimization

Every function maximizes its own metric. The result: nobody owns the enterprise number.

03

Single-future plans

One number stands in for an uncertain future — variability arrives after commitment, on a margin you can’t recover.

04

Stale master data

Parameters describe a world weeks out of date. The world they model has already moved on.

Naming the move

A proposal is not a decision.

A planning engine produces a proposal: optimal only if one future is true. Deciding under uncertainty, across goals in conflict, is a different move — and the gap between the two is the whole reason the decision layer exists.

An order proposal
  • Assumes one future is true
  • Optimizes one objective — or none
  • Re-plans from scratch each cycle
  • Asks: how do we meet the forecast?
  • Measured by plan attainment
A decision
  • Weighs the full range of futures
  • Balances objectives in tension — service · margin · cash · risk
  • Acts now or waits — by the value of waiting
  • Asks: what should we do, given variability and risk?
  • Measured by economic outcome
Resilience first

A faster wrong answer is still wrong.

You can’t automate a plan you were never able to trust. Run a brittle, single-future plan faster and you compound the churn — more expedites, more safety stock, automatically. Make the decision resilient first, and only then are speed and autonomy safe to scale.