hosted·ai Raises $19M Seed Round to Transform AI Infrastructure Economics

New Capital to Solve the GPU Waste Problem

hosted·ai announced the successful closure of a $19 million seed funding round led by Creandum on March 19, 2026. This San Jose based startup aims to radically transform the economics of AI infrastructure by addressing the systemic waste in current GPU deployments. Joining the round are Repeat VC, People Ventures, Z21 Ventures, Golden Sparrow, Hersir Ventures, and Tekton.

The Crisis of Underutilized Compute

Current GPU infrastructure suffers from significant inefficiencies, where static rentals result in roughly 60% of capacity sitting idle. Because traditional cloud models require customers to rent fixed instances for peak workloads, service providers face massive upfront capital expenditures that threaten profitability. This scarcity is particularly acute as the AI industry shifts from massive model training to localized, low-latency inference.

Scaling Utilization by 5x

The company provides a hardware-agnostic software stack that enables GPU pooling and optimized multi-tenant workload placement. This approach allows providers to increase hardware utilization by up to 5x, effectively removing the high entry barriers for regional neoclouds. By transitioning to a consumption-based model, developers can finally pay for actual usage rather than reserved hardware.

"The GPU market has a waste problem, not a scarcity problem," said Ditlev Bredahl, CEO of hosted.ai. "We are building the operating system for the GPU economy, and this round puts us in a strong position to do exactly that."

Roadmap for the GPU Economy

Moving forward, the company is developing a unique GPU Mesh resource exchange to facilitate the buying and selling of spare capacity. This roadmap also includes the expansion of packet.ai, a developer-centric neocloud, and GPUaaS.com, a wholesale matchmaking service. These tools will enable service providers to launch full neocloud services without additional hardware investment. The founding team brings deep expertise from scaling UK2 Group and founding OnApp. Their history with VMware, Nvidia, and XenSource provides the technical foundation needed to navigate the complex GPU landscape. Based in San Jose, the company is now positioned to lead the transition toward a more sustainable and sovereign AI infrastructure.