Fourteen years building the analysis behind planning decisions — resilience, demand quality, integrated business planning. Reproducible, from first principles. Every number computed, not asserted.
02/05
How the work gets done. Written down so it can be held against every project on this page.
Judgment over code craft.
The hard part of planning is not the pipeline — it is deciding what the pipeline should prove.
Every number computed, not asserted.
If a figure appears in this work, a script produced it — and can produce it again on demand.
Reproducible from first principles.
Seed everything. Same inputs, same answer, every run. An analysis you cannot rerun is an anecdote.
Bounded claims, labelled data.
A model that cannot say where it stops working cannot be trusted where it does. Synthetic data is always labelled synthetic.
The forecast is a decision,not a prophecy.
03/05
Which node, if it fails, hurts most — and for how long?
Time-to-Recover / Time-to-Survive stress testing in the Simchi-Levi tradition. Every node in a synthetic multi-echelon network is failed in turn; the model measures how long downstream demand survives on inventory and how long recovery takes. The nodes that hurt are rarely the ones the spend report points at. Seeded and reproducible — same network, same answer.
View repositoryIs our forecast actually adding value — and where is it worst?
Forecast Value Added in the Gilliland tradition: every touch in the forecasting process measured against a naïve baseline, so effort that subtracts accuracy is named, not defended. Crossed with ADI/CV² demand segmentation (Syntetos–Boylan–Croston) to show where the demand signal can reward attention at all — and where the honest answer is a simpler model.
View repositoryBase vs upside vs constrained — what do we commit, and what does it cost?
S&OP / IBP scenario reconciliation with rough-cut capacity checks, in the Wallace / Oliver Wight tradition. Three demand scenarios pushed through the same supply model; the gaps, the capacity breaks, and the cost of each commitment computed side by side. The arithmetic behind the executive S&OP meeting — done before the meeting, not asserted in it.
View repositoryAll demo data synthetic & labelled · seeded · reproducible
04/05
0
Years in planning
Functional and delivery, across demand, supply and executive S&OP.
0
Planning platforms
SAP IBP · o9 · Kinaxis · OMP · SAP APO.
0
Industries deep
Pharma cold chain, FMCG, energy — depth over logo count.
0
Programs delivered
With a global consulting firm, Big Four scale.
05/05
Available for the hard planning questions — the ones where the answer has to survive scrutiny.