{"slug":"shaping-global-infrastructure-with-ironclad-manufacturing","url":"https://findausecase.com/use-cases/shaping-global-infrastructure-with-ironclad-manufacturing","title":"Shaping global infrastructure with ironclad manufacturing","description":"Gerdau, Brazil's largest steel producer, consolidated a complex homegrown open-source data ecosystem onto the Databricks Data + AI Platform, using Delta Lake, Delta Sharing and Unity Catalog to enable real-time processing for digital twin simulations of production lines. The move cut data processing time from 1.5 hours to 12 minutes, delivered a 40% cost reduction for data processing and 80% faster development of streaming solutions, and enabled new AI applications including predictive maintenance, image and text classification, and an LLM-based upskilling assistant for employees.","company":"Gerdau","industry":"Manufacturing","country":"Brazil","aiCapabilities":["Digital Twins & Simulation","Predictive Analytics","Computer Vision","Large Language Models"],"technology":["Databricks Data + AI Platform","Delta Lake","Delta Sharing","Agent Bricks","Unity Catalog","Photon","Power BI"],"deployment":"Public Cloud","problemStatement":"Gerdau's homegrown, open-source-based data tools were complex, disconnected and hard to manage, requiring users to be proficient in Python and Spark, which made it difficult for engineering to maintain and drive adoption. The platform also lacked real-time data processing, hindering the company's digital twins use case, and lacked fine-grained access and data lineage controls needed for compliance and security as the company scaled.","solutionApproach":"Gerdau consolidated its data tools onto the Databricks Data + AI Platform, with Delta Lake as the underlying data infrastructure and Delta Sharing to securely share data internally and with manufacturing and distribution partners. Using Photon, the platform's next-generation query engine, Gerdau sped up data processing, and Unity Catalog established data governance standards with fine-grained access controls, data lineage controls and access segregation, paired with Power BI for business reporting. The company also began exploring digital twins, predictive maintenance, image and text classification, and a generative-AI-powered LLM assistant to help employees with re/upskilling.","businessValue":"Gerdau achieved a 40% cost reduction for data processing and an 80% improvement in new developments for streaming solutions, reduced average data processing time from 1.5 hours to 12 minutes, cut ML model creation effort by 30%, and rapidly onboarded over 300 new global data users across countries including Peru and the United States.","evidence":{"band":"high"},"sourceUrl":"https://www.databricks.com/customers/gerdau","dates":{"publishedAt":"2026-09-21T05:46:53.333Z","publishedAtSource":"pipeline","updatedAt":"2026-09-21T05:46:53.333Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/shaping-global-infrastructure-with-ironclad-manufacturing. Bulk republication requires permission."}