Multi-site rollout
A repeatable deployment pattern across sites and regions, with site-level autonomy settings and group-level reporting.
One line proves it. One factory operationalises it. An enterprise agreement makes Duromex the way your group builds appliances — with custom models, a managed edge fleet, contractual SLAs and outcome-based terms.
Built for multi-site appliance groups
Design-partner programme, 2026. Duromex is pre-launch — partner identities are withheld under NDA and shown by segment, not by logo.
Everything here exists because a group standardising on Duromex has problems a single plant does not.
A repeatable deployment pattern across sites and regions, with site-level autonomy settings and group-level reporting.
Dedicated product, line and quality models trained on your platforms rather than a shared baseline.
A managed factory-edge GPU fleet with versioning, canary rollout, rollback and OTA updates across every plant.
Per-tenant scoping end to end, strict recipe IP protection, and a fully on-premises option for sensitive producers.
Contractual uptime, response and escalation commitments with a named engineering contact.
A share of measurable first-pass-yield, warranty, scrap and throughput improvement, for the workflows where value is unambiguous.
One plant, one line, one metric, in shadow then advisory. The reference that every subsequent site is sold against internally.
Adjacent modules in the same plant, then the full loop, then the second plant on the same pattern.
A managed edge fleet across sites, with model rollout, canary and rollback handled centrally.
Duromex becomes the system of action: the audit trail of record, the recipe home and the encoded craft of the group.
Price follows value, and the term follows the proof.
| Component | How it works | Why |
|---|---|---|
| Platform subscription | Tiered by line and factory across the group | Predictable base that scales with footprint |
| Usage | Per-unit, per-line or per-action metering above plan limits | Aligns cost with actual production volume |
| Outcome component | A share of measured first-pass-yield, warranty, scrap or throughput improvement | Puts our margin behind our claim |
| Add-ons | Dedicated models, on-prem, premium support, custom integrations | Depth where the group needs it |
| Annual prepay | 15–20% discount | Improves cash and retention on both sides |
| Multi-year | Usage ramps with committed expansion | Discounts traded for term length and case studies |
In a multi-site group, autonomy is a governance problem before it is a technical one. Duromex models authority explicitly: which agent, on which line, at which site, may write what — and who approves a change to that.
Every grant is revocable, every action is logged immutably, and the twin gates any change to a validated process.
modelled lifetime value at 135% NRR
blended enterprise CAC
target payback period
target gross margin at scale
Modelled unit economics from the Duromex financial model. Pre-launch figures, not audited results.
Every enterprise buying committee raises the same four. None of them are unreasonable.
Shadow mode with a published accuracy gap against your own baseline, before any write access is discussed.
Human checkpoints on high-impact decisions, fail-safe stop paths, twin validation and an immutable audit trail.
Graduated autonomy means the line changes at the speed the team accepts, not the speed the software allows.
Pre-built connectors for forming, paint, foaming, test, assembly and MES, with a deliberately narrow first workflow.
The reason multi-site works is that the loop is identical. What changes is the product, the equipment vendor and the local constraints — all of which are configuration, not a rebuild.
Enterprise agreements make these contractual.
uptime target
named engineering contact per group
quarterly ROI reviews per year
actions retained in the audit trail
“Our end-of-line functional testers told us what failed. They never told us which press stroke or foam shot caused it. That link is the whole product.”
“Foam voids are invisible until the energy rating comes back wrong. Predicting density distribution before the cabinet cures is the part I could not buy anywhere else.”
“If it writes to the press, it needs an audit trail a quality auditor accepts. Duromex started there instead of bolting it on.”
Yes. An on-prem option exists for sensitive producers. Cross-site learning is federated, so shared models improve without your recipes or quality data crossing a boundary.
Your data and, in enterprise agreements, model artefacts trained on it remain yours and are exportable. We would rather retention came from the loop working than from hostage-taking.
Against the baseline established in shadow mode, on a metric agreed in writing before the pilot, with the measurement method documented and auditable in the platform.
Yes — through regional deployment and the on-prem option, with per-tenant scoping enforced throughout.
Start at one plant, on one line, with one number. Standardise from the proof.
Duromex is pre-launch. Figures shown are design-partner targets and modelled economics, not audited results. Ask us for the methodology.