Resilient.
From best-mean plans to policies that hold under uncertainty.
Plans break. Policies hold. Every committed decision gets a resilience score before it commits — the percentage of plausible futures in which it remains the right call. Tail cost gets priced into the objective, not noticed in the post-mortem.
Best-mean plans break the moment reality drifts.
Enterprise Resource Planning (ERP) carries a single lead-time number per part. Reality carries a distribution — 9 days when the supplier is on rhythm, 31 days when the port is congested. The plan averages the two and protects neither.
Risk modeling, where it exists at all, is binary: did the scenario happen, yes or no. The dollarized probability curve — what is the cost of this shock, weighted by the likelihood it materializes? — never enters the math.
Best-mean optimisers maximise expected outcome and leave the tail outside the objective. The tail materialises once a year on average, and the post-mortem that follows tends to read like the post-mortem from three years ago, because the architecture that produced it has not changed.
What it costs.
Working capital trapped in uniform safety stock that nobody can defend on a balance-sheet review.
Strategic-account OTIF breached during constrained periods. Future revenue silently shifts elsewhere.
Named risks with a dollarized probability curve. Mitigation premium decisions made on gut, not ROI.
The worst-case scenarios that materialize once a year are never represented in the planning objective.
What we bring.
- 01Distribution-driven inputs
Every driver — lead time, scrap, yield, demand, cost — is a learned shape from history, not a value.
- 02Iteration ensemble
Thousands of futures sampled by Monte Carlo or Latin Hypercube. The plan sees the full uncertainty cone before it commits.
- 03Outcome constraints
A service floor, a margin floor, and a carbon ceiling each become either hard or soft bounds on the enterprise objective, depending on how the policy is configured.
- 04Resilient EVA + CVaR
VYAN's objective combines Resilient Economic Value Add () with a Conditional Value at Risk () tail penalty — mean expected enterprise value minus a dollar penalty on the worst-tail outcomes. The objective explicitly trades expected outcome against tail fragility.
- 05Decision Resilience Scoring
Every committed decision carries a forward score: in what percentage of plausible futures does this decision remain the right call?
- 06Named risk events with probability curves
Port closures, supplier failures, demand pull-ins — each carries a probability distribution and dollarized impact.
How it lands on the P&L.
Stress-testing belongs before commit, not after the shock has already cost you.