hosted·ai has closed a $19M seed funding round to accelerate its mission: making AI infrastructure simple, efficient and affordable for service providers and their developer and enterprise customers. The round was led by Creandum, with Repeat VC following, and participation from existing investors People Ventures, Z21 Ventures, Golden Sparrow, Hersir Ventures and Tekton.
The problem: GPU infrastructure is broken
AI depends on GPU infrastructure – but today’s GPU is profoundly wasteful. Unlike traditional cloud compute, which scales dynamically to match demand, GPUs are static: customers must rent fixed instances based on estimated peak workload requirements. This creates three compounding problems:
- CAPEX and profitability: service providers operating GPU infrastructure (“neoclouds”) must make enormous upfront investments to meet customer demand, making profitability a persistent challenge.
- Utilization and waste: AI workloads consume only around 40% of GPU capacity on average, meaning roughly 60% of the capacity that neoclouds invest in – and customers pay for – sits idle.
- Scarcity and access: as AI shifts from model training to inference, companies need local, low-latency, sovereign GPU infrastructure. High CAPEX requirements and GPU scarcity outside hyperscalers make it extremely difficult for regional service providers to meet that demand.
The solution: the hosted·ai software stack
The hosted·ai software stack tackles these problems directly, transforming the way GPUs are orchestrated, managed, sold and used:
- hosted·ai: our core GPUaaS software platform. Through GPU pooling, optimized multi-tenant workload placement and GPU overcommit, hosted·ai delivers up to a 5x improvement in GPU utilization – reducing CAPEX requirements 5x, boosting profitability, and removing entry barriers that have kept regional service providers on the sidelines. Next on the roadmap: GPU Mesh, a unique resource exchange that enables service providers to buy and sell spare GPU capacity, extending their reach – or launching a full neocloud service – without additional hardware CAPEX.
- packet·ai: a neocloud powered by our customers’ optimized GPU infrastructure, delivering GPUaaS at market-leading prices. packet.ai generates direct demand for hosted·ai customer GPUs, and forms the foundation of an open-source neocloud portal that customers can deploy at launch to accelerate their time to market.
- GPUaaS.com: a wholesale GPU matchmaking service connecting enterprise buyers with hosted·ai customers and partners that can fulfil custom GPU cluster requirements at scale.
Ditlev Bredahl, CEO of Hosted.ai said: “The GPU market has a waste problem, not a scarcity problem. We’ve spent 25 years building infrastructure software that makes service providers competitive – and the GPU opportunity is the biggest we’ve seen. This funding lets us move faster: more platform, more partners, more regions. We’re building the operating system for the GPU economy, and this round puts us in a strong position to do exactly that.”
hosted·ai was founded by a team with deep experience of infrastructure technologies and service provider ecosystems. Ditlev Bredahl, Narendar Shankar, Julian Chesterfield and James Withall were previously instrumental in scaling UK2 Group, and founding OnApp, to bring cloud infrastructure as a service to the mainstream service provider market. They have also led numerous infrastructure technology, market development, AI and ecosystem initiatives at companies including VMware, Expedia, XenSource and NVIDIA.
