Fourteen years inside supply-chain planning — SAP IBP, o9, Kinaxis, OMP, S&OP — most of them at a Big Four consulting practice. Pharma cold chain. FMCG. Energy.
I build the analysis behind planning decisions: resilience, demand quality, integrated business planning. From first principles, reproducibly, and increasingly with applied AI in the workflow. Every claim on this page is a working method, not a slogan.
Run Forecast Value Added: score every human touch against a naïve baseline. Wherever the touched forecast loses to naïve, the touch is the problem — and it gets named, per planner, per SKU. Method: FVA — Gilliland
A forecast allocates stock, capacity and cash. Judge it by the cost of being wrong in each direction — not by how comfortable the single number feels in a review meeting. Consequence: bias has a price tag
Seeded runs, versioned inputs, printed assumptions. Same inputs, same answer — or the number does not ship. Synthetic data is labelled synthetic, every time. Rule: reproducible or rejected
Time-to-Recover versus Time-to-Survive, node by node. Where TTR exceeds TTS, you have found the failure before it finds you — and you can price the fix. Method: TTR/TTS — Simchi-Levi
The hard part of planning is deciding what to compute and what being wrong costs. Code is the cheap part — which is exactly why it should be boring, seeded and checkable. Corollary: boring code, sharp questions
“Which node, if it fails, hurts most — and for how long?”
METHOD → Time-to-Recover / Time-to-Survive stress testing, in the Simchi-Levi tradition. Every node failed on purpose; impact reported in days and margin, not adjectives.
VIEW REPO ↗“Is our forecast actually adding value — and where is it worst?”
METHOD → Forecast Value Added (Gilliland) plus ADI/CV² demand segmentation (Syntetos–Boylan–Croston). The touches that hurt get named; the demand that can't be forecast gets planned differently.
VIEW REPO ↗“Base vs upside vs constrained — what do we commit, and what does it cost?”
METHOD → S&OP/IBP scenario reconciliation with rough-cut capacity, after Wallace and Oliver Wight. One set of numbers the business can sign — with the cost of each alternative printed beside it.
VIEW REPO ↗