When downtime costs six figures an hour, service is the product.
Fabs expect near-perfect uptime and rapid parts response. An AI agent command centre triages every tool alert across the fleet and turns it into dispatch-ready, priced action.
Two verticals. One platform.
Fab tool makers and hyperscaler data-centre builders share one hard problem: the asset you install is the asset you service for a decade. ServiceCPQ runs both on one knowledge graph, so nothing is re-entered between build and operations.
Keep fabs running. Turn service into revenue.
For process-tool makers, a decade of service, parts and contract revenue rides on each install. ServiceCPQ turns that installed base into a governed, quotable revenue stream.
- Installed base per customer, tool and process node
- Coverage-gap detection with one-click auto-quote
- Quote → contract → invoice on one workflow
- Usage-based, flat-fee and T&M pricing
- Depot repair and loaner-pool tracking
- AI agent triage → priced dispatch
Build the rack. Own the asset. No re-entry.
An ERP treats a data-centre build like an installation order — the same single object you’d raise to wire an AC unit. A hyperscaler build is a project: milestone completion, milestone billing, retention and T&M capture — the same shape as underground mining-equipment installation. Force it into one install order and you end up with a stitched, multi-module workaround; ServiceCPQ models it natively as a project, then hands the finished rack to the Work Order Copilot as a live asset.
- Rack position grid — zone, hall, rack, slot
- Milestone billing gates on completion
- T&M cost capture against the build
- Invoice draft with gross margin
- 5% retention tracked to close-out
- Knowledge graph: zone → rack → GPU slot
Service is judged in minutes, not days.
Unforgiving uptime SLAs
A missed intervention window is a contractual and reputational event, not just a cost.
Reactive, not predictive
Rich tool telemetry exists but rarely triggers a service action before failure.
Complex contracts, slow quotes
Multi-tier service agreements take weeks to configure and price.
Traceability demands
Every part and intervention must be logged against the exact tool and process.
What's really at stake in fab uptime.
Industry benchmarks shown as context, not ServiceCPQ results.
Built for the way fab equipment is serviced.
Asset health, entitlement and complex-contract quoting on one model — so uptime commitments are backed by the system, not spreadsheets.
Agents triage tool alerts across the fab, rank priority and recommend the next action from one console.
Technicians work a live, ranked feed of flagged tools with the fix and part pre-attached.
Every fault and condition alert becomes a priced, dispatch-ready service opportunity.
High-value module life tracked per tool, triggering rebuild and exchange quotes before failure.
The position ID that ends the handover problem.
When a rack is installed, its build-time position — zone, hall, rack, slot — becomes a permanent node in the knowledge graph. Operations, PM, SLA and billing all reference that same node. No CSV export, no re-keying, no reconciliation between the build system and the service system.
Where ServiceCPQ changes the economics.
Zero re-entry asset handover
A finished rack is re-keyed from the build tracker into the service system by hand — weeks of reconciliation, and errors that surface as the first SLA miss.
The install-time position ID is already the asset node. The day a rack is energised it is a live, serviceable asset — nothing re-entered.
Milestone billing gate enforcement
Invoices go out before a milestone is actually signed off, or lag weeks behind completion because billing lives in a different system than the work.
Billing gates sit on the milestone itself — the invoice drafts the moment completion is verified, with retention and T&M already applied.
Coverage gap → quote in one click
A tool runs out of warranty or contract coverage and no one notices until months of billable service have been given away for free.
The graph flags the coverage gap the day it opens and drafts the renewal or T&M quote automatically, ready to send.
Fits your existing stack.
ServiceCPQ sits over your systems of record, not instead of them — the graph orchestrates build, service and billing while your ERP stays the ledger. Native SAP S/4HANA and BTP connectors mean it drops into a hyperscaler or fab-OEM landscape without a rip-and-replace.
See it running on a real DC build programme.
Virginia DC-1 — a $18.5M hyperscaler build — modelled end to end: milestone billing gates, 5% retention, T&M capture, and every rack handed to operations as a live asset the day it’s energised.