Geminus

Geminus is an industrial AI platform for optimizing complex physical systems such as refineries, gas production sites and electric grids. Instead of learning patterns only from historical data, it builds models that start from the underlying physics of an asset and combines simulators, historical data, real time sensor feeds and operational context. Recommendations come with uncertainty qualification so that operators can judge how far to trust them. The product page describes an AI generative engineering platform that gives engineers and data scientists self service tools. A user describes the problem, the platform contextualizes the data, and it trains surrogate models in hours, with training artifacts checked automatically and available for manual inspection. The surrogates and an optimizer can then be wrapped in an agent with a user interface that is shared with the operations team. Customers who prefer not to build themselves can use Geminus directly or certified builder partners. It is aimed at energy and industrial operators running high consequence operations where downtime or errors are costly. The vendor cites a US gas producer that cut methane flaring by 25 percent while raising production by 5 percent, a refiner that moved from pilot to production in days, and $100 million of value delivered in the first year. These are vendor reported results. Pricing is not published. The site offers a demo request and a contact option, and sales are handled directly.