modules
Planning and Intelligence
Stop arguing about the plan — rerun it and watch the constraint confess.
Planning fights have no referee: the spreadsheet says one thing, the floor says another, and the meeting decides by seniority. Deterministic solvers end the politics — same inputs, same plan, with an execution profile showing exactly which constraint binds. Disagree and you point at the constraint, change it, and rerun to watch the difference.
Why replayable plans end arguments
Planning fights usually have no referee: the spreadsheet says one thing, the floor says another, and the meeting decides by seniority. Deterministic solvers change the game — same inputs, same plan, with the execution profile showing exactly which constraints bound the outcome. Disagree with the plan and you can point at the constraint, change it, and rerun to see the difference.
data Viz
data View
section: thesis
Planning guarantees
- Deterministic runs
- Identical inputs reproduce identical plans with execution evidence
- Honest infeasibility
- Certificates naming binding constraints with sensitivity, never silent partials
- Live metrics
- Demand, adherence, throughput, and margin from operational state as it validates
- Versioned scenarios
- What-if branches cite their parent inputs and assumptions explicitly
- section
- capabilities
Who answers for Planning
Four roles, four anxieties, one deterministic record answering each.
data Viz
data View
section: personas
Section 1
col Title: COO
col Cont: Plans that defend themselves with execution profiles instead of planner seniority. Infeasible stretches confess their constraints with sensitivity attached.
Section 2
col Title: CFO
col Cont: Demand, margin, and utilization metrics read live state — forecasts anchor to facts the business already validated instead of models nobody can audit.
Section 3
col Title: Planner
col Cont: What-if branches version against live state without touching production plans. Promotion to active requires owner acceptance with evidence attached.
Section 4
col Title: CMO
col Cont: Demand sensing fuses orders, forecasts, and service signals with overrides versioned against baselines — bias stays visible instead of hiding in spreadsheets.
From spreadsheet politics to proven plans
subtitle: What each stage feels like after adoption.
layout: alternating
enable Slideshow: false
slideshow Interval: 5000
Signals unify
date: Phase 1
status: in-progress
Orders, forecasts, and service demand fuse with lineage intact.
has Learn More: false
Solvers prove
date: Phase 2
status: upcoming
Deterministic runs carry execution profiles; infeasibility confesses.
has Learn More: false
Metrics go live
date: Phase 3
status: upcoming
Demand, adherence, and margin read operational state without batches.
has Learn More: false
Scenarios version
date: Phase 4
status: upcoming
What-ifs branch safely and promote with owner acceptance on evidence.
has Learn More: false
Planning in daily operation
Each plan below runs deterministically against live state — same inputs, same plan, with the evidence to defend it and the honesty to admit infeasibility.
data Viz
data View
section: applications
Demand Plans
Orders, forecasts, and service signals fuse into demand with overrides versioned against statistical baselines. Accuracy reviews compare both lines so bias surfaces instead of hiding.
Replenishment
Purchase and transfer orders generate from demand with lot-size policy, lead times, and shortage causes named. Approved plans convert directly into executable supply without re-entry.
Transportation
Loads consolidate against carrier capacity and transit commitments. Late loads explain themselves with the constraint and the recovery option attached.
Labor Plans
Shift demand from production and service schedules resolves against the worker graph before rosters commit. Shortfalls name the missing skill so hiring targets reality.
Planning sub-modules
subtitle: Sense demand, plan the network.
data View: planning-intelligence:sub-modules
Demand Planning
desc: Plans with proof attached.
link: /sub-modules/demand-planning
Planning questions
Section 1
How fresh are planning metrics?
They read live operational state, so demand, utilization, and margin visibility update as work validates — no nightly batch window.
Section 2
What is an infeasibility certificate?
A solver output naming the binding constraints that prevent a feasible plan, with sensitivity showing how much relief each constraint needs. Planners adjust reality or expectations from the same document.
Section 3
Can planners test scenarios safely?
Yes. What-if branches version against live state without touching production plans, and promotion to active requires owner acceptance with evidence.
Section 4
Will this replace our planners?
No — it promotes them from spreadsheet mechanics to decision owners. Judgment moves to constraints and trade-offs while the solver handles arithmetic nobody should do by hand.
Section 5
Our demand is too volatile to forecast. Does this still help?
Volatility is exactly what versioned signals handle: overrides track against baselines with accuracy reviews, so the plan adapts weekly instead of pretending annually.
data View
section: faqs
content section
value: Every solver run produces an execution profile and evidence — infeasible plans explain themselves with sensitivity analysis instead of returning silent partial answers. Demand, fill rate, fulfillment latency, utilization, and margin stay visible without waiting for nightly batches.
Section 9
Demand sensing from orders, forecasts, and service signals
Section 10
Replenishment, transportation, and labor plan generation
Section 11
Deterministic solvers with execution profiles and evidence
Section 12
Infeasibility certificates with sensitivity analysis
Section 13
Live BI on demand, fill rate, latency, utilization, and margin
Section 14
Schedule adherence and throughput tracking without ETL
Section 15
What-if scenarios versioned against the same live state
Section 16
Plan-to-execution linkage with owner acceptance gates