Back in 2011, when we build OnApp, our mission was simple: help service providers compete against the hyper-scalers by offering easy-to-use cloud solutions with great UX and straightforward monetisation. We had an incredible run, capturing about 30% of the market in just a few years. Why? Because service providers excel at infrastructure, support, and uptime - they're great operators. But they aren’t always product-focused people. OnApp was all about simplifying cloud services and enabling these providers to build profitable, recurring revenue streams.

Before OnApp, during my time running UK2 Group, we transformed UK2 from a $1-2M mass-market host into a +$50M product-driven company, without external funding. With brands like VPS.NET and 100TB, we grew quickly without relying heavily on marketing. Our go-to-market strategy was clear: build products that were 5% better than anything else out there. This "we add 5%" mantra allowed us to significantly lower customer acquisition costs. Every dollar we spent on product development delivered 3x the return compared to spending the same on marketing for existing products.
It really is all about the product.

Now, I’m excited to share that I’m reuniting with the talented teams from UK2 and OnApp to build www.hosted.ai. So, yes - we're putting the band back together…Our goal? To do for AI what OnApp did for cloud. We’re creating a platform that makes consuming GPUs incredibly simple, removing concerns about GPU brands or generations. As a service provider, you'll be able to sell GPU resources just like CPU, RAM, or storage ... fully pay-as-you-go. Mix and match T4s with H100s, NVIDIA with AMD or Intel - it doesn’t matter.

And with hostedAI you can build high-density GPUaaS AI offerings, including overprovisioning. You can ensure profitability without compromising customer experience. It’s all supported by one-click deployments for training and inference optimised infrastructure.

In many ways, hostedAI mirrors what we accomplished in 2004 at UK2 and in 2011 with OnApp. We’re simplifying complex technology to empower service providers and help them build profitable, recurring revenue streams in emerging markets. And the market situation is also comparable to pre-onapp. The hyper-scalers (plus new comers like Lambda and CoreWeave) are once again threatening the service providers. It's like 2004 all over.

I’m really excited about this journey and the opportunities ahead. If you're a service provider looking to expand into AI and GPU services without complexities, let’s connect. Together, we can make AI infrastructure accessible, scalable, and profitable.

Feel free to reach out to me directly ([email protected] or DM) - I’d love to share more about what we’re building at hosted.AI


D