Ora Computing, a startup founded by two quantum computing researchers, has raised €3.5 million to build software that compresses and optimises AI foundation models, tackling the soaring compute costs of running large models at scale.
The seed round was led by Constructor Capital and Greencode Ventures, with continued participation from existing backers. The round type and specific valuation were not disclosed.
For enterprises operating at scale, compute costs can reach tens of millions of dollars annually, making efficiency a critical pain point. Ora's approach targets the optimisation layer between models and hardware, a space attracting increasing attention as AI deployment grows.
The funding signals investor appetite for infrastructure solutions that reduce AI's energy and cost burden. With hyperscalers and startups alike racing to deploy models, tools that compress models without sacrificing performance could become essential.
Founded by researchers with a background in quantum computing, the team brings deep technical expertise to a problem that many dismiss as secondary to raw model performance. "We aren't building another model. We are building the pulleys and levers that make existing models vastly cheaper to run," the founders noted.