The approach
PFEP as the foundation. MRP and Lean as the discipline that runs on top of it.
Plan For Every Part is the source of truth
Every planning, inventory, and MRP decision is only as good as the part-level data underneath it. Before touching buffer levels, replenishment logic, or Lean flow, the PFEP dataset has to be built and kept current — supplier lead times, pack sizes, usage rates, storage requirements, all structured against how the plant actually runs.
Most internal teams have pieces of this scattered across ERP exports, spreadsheets, and tribal knowledge. This engagement starts by consolidating that into one governed dataset.
Buffers and flow, not just data
Once PFEP is in place, DDMRP positions and sizes decoupling buffers against real demand variability — and Lean methodology governs how material actually flows through the plant. The two aren’t separate workstreams; DDMRP without a clean PFEP foundation is guesswork, and Lean without demand-driven buffers just moves the bottleneck.
A phased plan, not an open-ended engagement
Every engagement follows the same six-phase sequence, each gated on proof before the next begins — and each with a defined stopping point, so you can stop at the end of any phase and still own something useful:
- P0 — Mobilise & Charter: sponsor, scope, and baseline agreed before any data work starts. Exit gate: charter approved.
- P1 — Build the Data Substrate: a governed, refreshable PFEP record for every pilot part. Exit gate: data integrity at or above threshold.
- P2 — Instrument Variability & Segment: demand and lead-time variability captured; parts segmented by value and behaviour (ABC/XYZ). Exit gate: segmentation signed off.
- P3 — Position & Size Buffers: decoupling points selected, buffers sized, MRP policy set per segment. Exit gate: buffers sized and loaded.
- P4 — Pilot Execution: the pilot runs live against net-flow position with a daily status routine. Exit gate: pilot targets met.
- P5 — Stabilise, Govern & Scale: the cadence becomes permanent and rolls out beyond the pilot. Exit gate: cadence live, rollout approved.
Built on the plant floor, not in a slide deck
APICS CPIM certified, with forty years in precision manufacturing and supply chain operations, including senior roles at Integer Holdings and Parker Hannifin. This is Excel (Power Query, Power Pivot, DAX), Power BI, Oracle EBS, and JD Edwards experience applied directly — not theoretical.
AI-Assisted, Scaling With You
The PFEP dataset doesn’t stop being useful once the dashboards are built. A working AI agent — built on Claude, configured against the PFEP structure — turns the dataset into a daily analyst: DDMRP buffer-zone status (Green/Yellow/Red/Top-of-Green), stockout risk, excess and working-capital exposure, supplier concentration risk, and what-if scenario testing, run conversationally rather than through a fixed report template. This is a working system prompt and daily query library, not a concept — built and tested against real PFEP structures. It’s part of the standard engagement, not a separate line item.
The same approach extends to governance. A second configured agent applies Hoshin Kanri X-Matrix discipline to the PFEP programme itself — tracking whether phase-level objectives, targets, and improvement priorities stay aligned as the engagement moves from pilot into P5. Built for single-site operations first, it scales to a multi-site correlation view — cross-site comparison, best-practice signals, strategy-level drift — without switching tools, if and when an engagement grows in that direction. Worth being precise about what this does and doesn’t do: it surfaces misalignment and drift for a human to act on — it doesn’t run the catchball conversation or resolve the underlying alignment problem, which is still a leadership function, not a software one.
See how this applies to your operation.
The Services page breaks the P0–P5 structure into what’s actually delivered at each stage.
Supply Chain Data Insights 