JuliaHub has launched Dyad 3.0, an AI platform that uses autonomous agents to design and test physical systems. The software is designed to model complex machines, such as satellites and semiconductors, by connecting AI agents with physics simulations and scientific machine learning (SciML).

The company stated that the platform allows engineers to move from initial specifications to production control code within a single environment. This approach is intended to compress R&D cycles from months to days by allowing AI to reason about fluid dynamics, temperature, and other physical forces.

JuliaHub has also secured $65 million in a Series B funding round led by Dorilton Capital to expand its “Dyad” platform for hardware engineering. The funding round included participation from General Catalyst, AE Ventures, and former Snowflake CEO Bob Muglia.

“It’s not about helping engineers complete one small task at a time. It’s agentic engineering at scale, where teams can feed a full specification to Dyad and have it design the complete system,” said Viral Shah, chief executive of JuliaHub.

The platform is currently used by several Fortune 100 companies in the aerospace, automotive, and utility sectors. In one application with Binnies and Williams Grand Prix Technologies, the software used four sensor inputs to predict faults in water distribution pumps with more than 90% accuracy.