Nestlé USA

Across Nestlé USA's network of factories, spare parts data had fractured along plant boundaries. Rather than using manufacturer part numbers as a common reference, each facility assigned its own SAP material numbers — meaning the same component could be stocked under 20 different identifiers across 20 factories, with further duplicates inside individual plants. COVID-era supply disruptions had pushed inventory levels higher, and those buffers were never drawn down once supply normalized. Without cross-site visibility, technicians couldn't see stock at sister facilities, analysis of component usage across the network was unreliable, and teams had limited leverage in supplier negotiations. The operational cost was concrete: senior engineering staff fielded emergency search requests every Saturday morning, manually hunting parts across SAP.

Nestlé USA deployed SPARETECH's AI-enabled platform, integrated directly with SAP, to attack the data quality and workflow fragmentation simultaneously. The tool applies machine learning to detect duplicate parts in real time as a user begins entering a manufacturer part number — surfacing potential matches before a redundant record is created. AI also auto-generates standardized SAP descriptions, replacing the inconsistent free-text entries that had made search unreliable. For technicians on the plant floor, the interface functions like natural language search rather than SAP wildcard syntax, and catalog images allow visual part confirmation before making a storeroom trip. The system additionally flags discontinued components and future obsolescence dates. On the governance side, Nestlé replaced each plant's ad-hoc intake process — Forms, Power Query, paper sheets — with a single approved workflow requiring sign-off from both the storeroom supervisor and maintenance manager. A super-user-led rollout seeded adoption across the factory network.

Parts search time dropped approximately 50%, and the Saturday-morning escalation calls to senior engineering staff effectively stopped. Monthly tool utilization across the factory network exceeded 95%, monitored via Power BI dashboards reviewed on network calls with plant teams. Qualitative outcomes matched the metrics:

• Technicians can locate parts at sister facilities without corporate intervention
• The same component is no longer held under 20 separate material numbers, eliminating the duplicate stocking risk that obscured true network inventory
• Working capital reduction is now an active strategy: high-cost critical spares are being centralized at selected facilities rather than replicated at every plant, with transfers executed as needed

• Spare parts master data is a maintenance strategy, not an IT housekeeping task — standardizing on manufacturer part numbers as the network-wide reference is what makes cross-site search and visibility possible.
• AI-assisted duplicate detection only holds if governance accompanies it: defined approvers, a single intake workflow, and monitored compliance prevent new data debt from accumulating.
• Cross-site inventory transparency reframes stocking decisions from a per-plant problem into a network optimization problem, directly reducing working capital tied up in redundant safety stock.
• Tool simplicity is an adoption lever — if technicians can learn the interface in five minutes and it solves a daily frustration, compliance follows without mandates.
• Monitor adoption metrics in recurring operational reviews, not just at launch — sustained data quality requires ongoing accountability.