Announcing Tensor9’s $4M seed round

We founded Tensor9 to help SaaS vendors unlock enterprise customers that can’t share sensitive data. Today, I am excited to share that Tensor9 has raised a $4 million seed round led by top investors including Wing VC, with participation from Devang Sachdev and NVAngels, Level Up Ventures, and strategic angels with backgrounds at Google (Michael Stoppelman), Cockroach Labs (Spencer Kimball), Benchling (Amit Gupta), and Amazon (Gert Lanckriet). 

As an early engineer at AWS on the S3 team, I spent a lot of time helping enterprises understand what it meant to trust the cloud. Public cloud enables simplicity, scalability, and efficiency, but also introduces security issues and data movement costs. There is a clear trend towards a more hybrid model, where software runs where it needs to in the right environment. And that’s the underlying premise of Tensor9: we help vendors ship software instead of requiring their customers to ship data.

We are a team of early-AWS engineers with deep technical expertise that helped build the transition from traditional on-prem to the cloud. Now, we are building the next transition from cloud to hybrid models that give customers more control over their sensitive data.

Our founding team Michael Ten-Pow, Matt Michie, and Matt Shanker in Seattle, WA, USA.

We are seeing a strong and growing need for privacy and security in enterprise software, especially in regulated industries. SaaS vendors are facing significant challenges selling to high-security and regulated markets like healthcare, finance, bioinformatics, and government unless they are able to offer a new set of options. 

Historically, this has primarily been on-prem deployments, where the vendor’s app is deployed into the buyer’s physical data center. But more recently, customers are also seeking or accepting a cloud-prem model, where the vendor’s app is deployed into the buyer’s cloud account (e.g. AWS, GCP, Azure). This model (also known as BYOC; Bring-Your-Own-Cloud) provides the power and ease of public cloud with the security and data privacy aspects of on-prem. 

Cloud-prem has taken off because the cloud overall has matured; the idea that a bank would treat a cloud operated by Amazon or Microsoft to be their premises has now caught on, enabling vendors to sell to banks where the software is deployed in the bank’s AWS account. In addition, new AI architectures are demanding new deployment mechanisms. First, AI involves heavy data, where customer data is too large to practically move. But AI also introduces a new set of security concerns because AI-powered tools and automation require extensive system privileges.

While customers are pushing vendors towards these models, vendors are significantly challenged to keep up; both on-prem and cloud-prem models are expensive to build, maintain, and support. There is the cost of upfront engineering and managed services compatibility, but the ongoing operations is where things can really get ugly. It’s one thing to be able to deploy software into a bank, but it’s another to be able to find issues, debug them, and fix them, particularly when customers are often slow to roll out updates. Vendors often end up managing multiple products which can add significant expense and distraction.

Tensor9 solves for any-prem deployments: Enabling SaaS and AI vendors to deliver their existing SaaS products directly into customer-owned environments, whether cloud-prem, on-prem, or fully air-gapped, and across Kubernetes, virtual machine, or bare metal deployments.  This unlocks new markets for SaaS vendors without burdening the vendor with complex bespoke engineering and operations.

How does Tensor9 make this happen? In a nutshell, a vendor points Tensor9 at their infrastructure as code (like Terraform or CloudFormation) and their app is deployed into different customer environments, with each customer getting their own private stack. Tensor9 continuously syncs updates to customer environments, ensuring consistency. A digital twin architecture mirrors the deployment and operational state of a customer’s stack, which allows vendors to sync changes to ensure consistency across customers and environments, as well as sync logs, metrics, and hardware failures back to the digital twin – enabling vendors to observe, debug, and support customers as if they were using SaaS. 

For SaaS vendors, Tensor9 helps them unlock new customer accounts while focusing on their differentiated value. They can keep their existing stack and tooling, including AWS services such as S3, avoid upfront heavy engineering costs and time to market, and achieve lower ongoing support costs. Vendors build one product versus multiple diverging products.

For their customers, Tensor9 significantly reduces the burden to operate the software versus traditional on-prem software delivery mechanisms. Customers get security and change management controls to ensure they consume software safely.

So what’s next for Tensor9? We’ve been engaging deeply with customers across a swath of verticals including enterprise data management, enterprise search, agent platforms, and security. This new round of funding will allow us to continue to expand our team to meet the needs of these SaaS and AI vendors. This includes supporting more customer environments and vendor stacks, improving the vendor onboarding experience, and helping vendors manage customer stacks at scale. We’re excited to be on this journey, and deeply appreciate the support from our investors, advisors, and customers. 

If you’re a SaaS or AI vendor that needs to deliver your product into customer-owned environments, please reach out. We are excited to partner together to learn about your needs and help you unlock new business.