34th FVI Expert Breakfast

Topic: "The Battle Against Spare Part Chaos" – How ElringKlinger cleaned up master data and eliminated duplicates with Sparetech.

In this session, René Hackbart (ElringKlinger) and Lukas Biedermann (Sparetech) presented a best-practice case for global spare parts management.

• The duplicate problem: René reported that in the old SAP systems, the same component was sometimes entered three or four times under different numbers (e.g., once as "Festo Cylinder", once as "Cylinder 123"). This leads to unnecessarily high inventory levels ("Dead Capital").

• The Sparetech solution: Sparetech has a database with 40 million original manufacturer data. They match the poor customer data against these "Golden Records". Result at ElringKlinger: 3,400 duplicates deleted, 17,000 records corrected.

• 4 minutes instead of 17 minutes: Previously, creating a new spare part in SAP took 17 minutes (manual typing, error-prone). With the Sparetech app, it now only takes 4 minutes because all data (weight, tariff number, image) is automatically taken from the manufacturer.

• Cross-plant search: A huge advantage is transparency. A plant in China can now see if the required part is in stock at the plant in Germany. This saves expensive reorders and uses existing inventory ("Virtual Pooling").

• AI for short texts: Lukas showed a new feature where AI (Gemini) automatically generates clean SAP short texts (e.g., "Motor 5kW Siemens"), which are translated into 40 languages. This solves the problem of poor data quality at its root ("Shit in, Shit out").

Classification: ADAM works immediately – even in data chaos ("ADAM-First")

This episode shows a classic but laborious way of data cleansing. With ADAM, we take a more disruptive approach that brings the customer to the goal faster.

• Intelligence instead of pre-project: ElringKlinger had to clean up first to become efficient. The ADAM approach: We accept the chaos ("Brownfield"). Thanks to hybrid search (vector search), ADAM finds the component "Cylinder 123", even if the technician searches for "Festo Cyl.". The AI understands the context, so no manual cleanup is needed beforehand.

• Autonomous data refinement ("Self-Building Knowledge"): Instead of relying on an external database (like Sparetech), ADAM uses AI agents. These retrieve missing information (data sheets, images) directly from the manufacturer's site and store them in the customer's system. The advantage: The customer builds data sovereignty (the data belongs to them), instead of permanently "renting" it.

• Integration as a bonus, not a requirement: We do not position ourselves against Sparetech, but as independent. Do you have Sparetech? Great, we integrate the clean data. Don't have it? No problem. ADAM cleans and complements your data autonomously during operation. You don't have to wait to start.

Conclusion: Sparetech is a powerful tool for purchasing. ADAM is the solution for operational business, delivering immediate value without the customer having to invest years in master data cleansing beforehand.