Parasail says,

Cloud infrastructure is dominated by several big players in the industry: AWS, Azure of Microsoft and Google Cloud. While some may seem that AI is heading in a similar direction, the founders of Parasail believe that AI infrastructure will look very different – and are betting on the fate of their company on it.

Parasail He works with dozens of suppliers to deliver graphic processors on request for companies and businesses that want to build AI models and applications. Parasail provides customers with access to hardware, including H100, H100, H200, A100 of NVIDIA, with some of the costs that the current charge, according to the company.

“There are generally three cloud providers who manage the Internet and that’s how the Internet is restored when you look at AI,” said Tim Harris, one of Parasail’s co -founders and Swift Navigation CEO, before TechCrunch. “It is much more fragmented. Compute is much fungal and liquid, so you can actually let it go in a more horizontal nature, and we were really trying to do it.”

Harris added, “We didn’t want a world where AI was controlled from soup to hyperskalers.”

Harris and Parasail CEO Mike Henry had the idea of ​​launching a few years ago. Henry, the former GroQ CPO, told TechCrunch that he had spent a long time thinking about what would be needed to build AI infrastructure that could compete with NVIDIA. He said that when he realized that AI’s infrastructure was quickly built by many players, he saw the opportunity for a horizontal move.

According to Henry, the quick clip of innovation that happens in AI hardware made it difficult for companies to continue.

“We had to really focus on how to do (this) as simply for the client?” Henry said. “They are hardly up to date with the open source model editions.”

Harris and Henry started at the company in 2023, hired an engineering team and began to build in early 2024.

Today, there is no shortage of suppliers who want to help businesses and other companies build and scale their AI products. From hyperskalers such as NVIDIA and Microsoft to start -ups such as Together ai and Lepton aiCustomers have many options.

Parasayla’s founders do not think they are all the same.

Of course, all these suppliers provide graphic processors and AI infrastructure, Henry said, but believes that the patented technology that Paraselle has worked under the hood helps to stand out. This technology is what connects the Parasail graphics processors from various sources.

Wednesday marks the official launch of its platform, but Parasail is already working with dozens of customers, including Elicit, Teags & Biases and Rasa. The company also raised $ 10 million from a range of seeds in 2024, which participated in basic endeavors, thresholds, Buckley Ventures and Black Opal Ventures.

In order to acquire a meaningful market share, Parasail will have to go his head with hyperskalers and to rely on the search for graphic processors that continue to grow. There is reason to believe that it will do it, but there are also signs – as Microsoft cancels some of their contracts for the Data Center – that the forecasts surrounding the required AI infrastructure may be a little overworked.

“We see literally any end (until) demand,” Harris said. “It really is that customers are hard to do – it is difficult to scale AI. The models are already reaching a place where (companies) can simply grab models with open source and almost execute them, but then we can access graphic processes, access to data centers, all optimizations … We can do so by clicking a button.”