Enterprises want access to new software and AI tools but can’t risk sending their sensitive data out to a third-party software-as-a-service (SaaS) providers. Tensor9 looks to help software companies land more enterprise customers by helping them deploy their software directly into a customer’s tech stack.

Tensor9 converts a software vendor’s code into the format needed to deploy into their customer’s tech environment. Tensor9 then makes a digital twin of the deployed software, or a miniaturized model of the deployed software’s infrastructure, so Tensor9’s customers can monitor how the software is working in their customer’s environment. Tensor9 can help companies deploy into any premise ranging from cloud to bare metal servers.

Michael Ten-Pow, Tensor9’s co-founder and CEO, told TechCrunch that Tensor9’s ability to transfer software to any premise, and its use of digital twin technology to help with remote monitoring, helps Tensor9 stand out from other companies, like Octopus Deploy or Nuon, that also help companies deploy software into a customer’s environment.

“You can’t just throw a piece of software over the wall, or it’s very difficult to throw a piece of software over the wall, and know what’s going on, be able to find issues, debug them, fix them,” Ten-Pow (pictured above, left) said. “They see it running, they can debug it, they can log in and understand what the issues are and fix them.”

He said the timing is right for Tensor9’s tech due to tailwinds from the rise of AI. Enterprises and financial institutions want to adopt AI tech, but can’t risk sending their data to a third-party.

“An enterprise search vendor might go to, let’s say, J.P. Morgan and say, ‘hey, I need access to all your six parabytes of data to build an intelligent search layer on top of it so that your internal employees can have a conversation with their company’s data,’ there’s no way that’s going to work,” Ten-Pow said.

Ten-Pow, an ex-engineer at AWS, said he had a “long, fairly winding path” to the launch of Tensor9. He got the idea for the company while working on another potential idea that didn’t work out. He spent some time figuring out if he could find a way to make it easier for software vendors to get SOC 2 certified, a cybersecurity compliance framework, to help them unlock customers that required their vendors to have it.

While that didn’t work out, he discovered from customer calls that what enterprises really wanted was the software to just run in their own tech environment. But many software companies, especially startups, don’t have the resources to offer a bespoke on-premise option for each enterprise customer.

That sentiment became the basis for Tensor9 which Ten-Pow launched in 2024. Later in the year, he brought on two of his ex-AWS colleagues, Matthew Michie and Matthew Shanker, as co-founders.

The company found early traction with voice AI companies. Since then, they’ve started to expand to work in other verticals including: enterprise search, enterprise databases and data management. The company currently works with AI companies including: 11x, Retell AI and Dyna AI, among others.

Tensor9 bootstrapped for its first year and recently raised a $4 million seed round led by Wing VC with participation from Level Up Ventures, Devang Sachdev of Model Ventures, NVAngels, an angel group of ex-Nvidia employees, and other angel investors. Getting investors on board with the idea wasn’t too challenging, Ten-Pow said, because the VCs they spoke with had seen their portfolio companies struggle with this exact problem. Tensor9 just had to convince investors that they were the right team for the job.

“We have a simple model but underneath the covers there’s a lot of complexity that makes that happen, hard technical challenges that we’ve solved to make that happen,” Ten-Pow said. “I think that was one of the things that helped us convince the investors to invest in us.”

The company plans to use the funding for hiring and for building out the next generation of its technology so that it can work with customers in more verticals.

“There’s been an evolution from [on premise] to the cloud and we think that this idea of software lives where it needs to, and operates where it needs to, is that next step that’s a sort of synthesis of the previous on-premise and cloud ideas,” Ten-Pow said.