About
Every container is a promise
Signed by hand — on a page the machine helped build, the one mark that can't be regenerated.
Functional and delivery, across SAP IBP, o9 and OMP. I care about the unglamorous part: whether the forecast is trusted, whether the S&OP meeting decides anything, whether the safety stock is sized for the service level you actually promised.
Fourteen years inside supply-chain transformations — the kind where the planning system and the planning process both have to change at once.
Fourteen years in supply-chain planning, currently leading planning transformations as a consulting director. Always up for talking shop — process design, IBP / o9, S&OP cadence. Views here are my own.
One throughline, five ways in.
The MBA years taught the vocabulary — supply chain as a system of tradeoffs, not a checklist. The fourteen years since have been spent testing that vocabulary against real forecasts, real safety stock, real S&OP meetings that either changed something or didn't.
Two other things run in parallel: zuloma.com and automationdiary.com, both built and kept running the same way this site is — as working proof, not a portfolio slide. Same instinct wherever it shows up: don't describe the system, ship the system, then explain what it does.
Travel gets planned the way a network gets designed — deliberately, in advance, city by city — not chased on impulse. That same disposition (plan first, take the calculated risk, protect the downside) runs underneath the SCM work too: safety stock exists because someone already decided what failure is allowed to cost.
This site is the newest experiment: built end to end with Claude Code, in the open, updated at the same pace as the work, the books, and the travel that feed it. Not a claim of AI fluency — the build log is the proof, page by page.
What the meetings taught me.
A forecast isn't finished when it's published. It's finished when the plant, the sales team and the planner argue from the same number.
If the S&OP meeting ends without a decision, it was a status update with better slides.
Service level is a promise with a price tag. I size the buffer against the promise, not against whatever survives the budget round.
A planning-tool rollout without the meeting redesign is redecoration. Both change in the same quarter, or neither does.
I'd rather size your safety stock than draw your roadmap. The unglamorous part is where the money lives.
Three places this shows up.
Work
Fourteen years of S&OP, demand planning and safety-stock work — case studies and two live planning calculators.
The proof, case by case→02Library
112 books across 11 shelves — read, reviewed, and opened into interactive decks.
Pull something off the shelf→03Lab
AI experiments and open-source repos — a working twin chat, a build log, real GitHub work.
See what's on the bench→If any of this landed near a problem you own.
Fourteen years of this leaves opinions. Most are already on this site; the rest come out by email.
Talk shop means the forecast nobody trusts, the S&OP meeting that decides nothing, the IBP-or-o9 call you're about to pay for twice. Plain email — no form, no calendar link.
Views here are my own. The safety-stock mistakes were too.