CoreWeave has launched a new service called Physical AI Field Engineering to assist enterprise teams in integrating AI into industrial workflows. The company has hired specialized engineers with backgrounds in automotive, aerospace, and mechanical engineering to work directly with client teams.
The process begins with a workshop to evaluate workflows and identify use cases for AI. CoreWeave claims this approach can reduce testing times by 17% to 35%. The service utilizes the company's cloud infrastructure, including bare-metal servers and tools such as Weave, marimo, and ARIA.
CoreWeave has already completed over 100 engagements with early adopters. For example, Nissan Motor Co. used the service to analyze 90 years of archived test data, which reduced chassis bolt-joint testing time by 17%. Another automaker completed an engine calibration task in 24 hours instead of the usual three months.
The new service is a result of CoreWeave's acquisition of the startup Monolith AI Ltd. in September 2025. The initiative aims to bridge the gap between industrial domain expertise and applied machine learning by using the customer's own data for model training.
