Why do hardware engineers need an "operating system" when they already drown in tools? In this session recorded at the Tech for Industry Show, Stéphane and Valentin (Koddex) answer with a key concept for the factory of the future: the Engineering Operating System.
The starting point: an industrial product means tens of thousands of components, thousands of specifications and hundreds of regulatory requirements, scattered across PLM, ERP, QMS and a maze of Excel files — to the point where half of a project's cost goes into checking data, tracking validations and handling non-conformities. The answer: anchoring three layers together — business processes, data models and operational data — through semantic links (component ↔ requirement ↔ test case) to build a true enterprise knowledge graph. The payoff: instant change impact analysis, native traceability across the entire product lifecycle, lower alignment costs, and structured data that is ready for agentic AI — fewer hallucinations, fewer wasted tokens. All illustrated with a 4-step implementation method, the "data compiler" analogy and the Koddex platform.
An essential watch for engineering leaders, design offices, PLM and quality managers who want to scale engineering rigor and get their data ready for the age of AI agents
