Startup Groundcover joined the crowded observability market with a few differentiators: eBPF-enabled visibility into application stacks, an architecture that keeps data in the customer's environment, and a node-based rather than volume-based pricing model.
When Groundcover entered the observability market in 2021, it was already a mature market with established vendors. But co-founder and CEO Shahar Azulay, who was working as a machine learning manager at Apple at the time, saw gaps in the capabilities of existing observability tools and challenges that practitioners like himself had to figure out on their own.
Groundcover was created to address two fundamental problems in the observability space: the complexity of instrumenting applications, and the unpredictable costs associated with data volume-based pricing models. These challenges often force companies to monitor less of their infrastructure than they’d like to, creating blind spots in critical systems.
As it turns out, Groundcover’s founders weren’t the only ones with those observability pain points. The company has grown since its launch, and this week Groundcover announced it has raised $35 million in a Series B round of funding, bringing total funding to date for the startup to $60 million.
“I’ve always managed R&D teams of some form, and we always use observability solutions,” Azulay told Network World. “So Groundcover combines these two things – firsthand experience with something that is painful, that was very mission critical to me and to everyone I worked with throughout my career. But also something very deep, … innovative, close to my core competencies.”
Open-source technology at the core: How eBPF changes observability
At the core of Groundcover’s platform is eBPF (extended Berkeley Packet Filter) technology. The open-source eBPF technology is part of the Linux kernel and has been widely used for network observability over the years, with technologies like Cilium (developed by Isovalent, which was acquired by Cisco in 2024).
In traditional network applications, eBPF allows developers to inspect network packets in real time, create sophisticated network filtering rules, and monitor network performance at the kernel level.
Groundcover’s expanded eBPF approach goes beyond traditional network monitoring, Azulay said: “eBPF is no longer just about network monitoring. We use it as an x-ray into operations flowing through the kernel of the operating system.”
Groundcover uses eBPF to provide full application-level traces. Azulay explained that the system can see the payload of the request and the response. He noted that, for example, if a user is sending a prompt to OpenAI, Groundcover’s eBPF-based technology will see the exact prompt and the response from the server.
How Groundcover uses eBPF to enable observability
Groundcover uses two types of eBPF probes to get information. The first type is kernel space, which gets information coming directly into the Linux kernel at an operating system level. The second is via user space probes, which gives visibility into running applications.
“We can actually observe things happening in the application stack,” Azulay explained. “Even if a packet is encrypted when it goes through the Linux kernel, we know how to observe it before encryption in the actual application stack.”
Unlike traditional network-focused eBPF tools, he claimed that Groundcover’s approach provides full application monitoring, performance profiling and detailed transaction insights.
“eBPF kind of replaces a lot of the things we expect” from an application performance monitoring (APM) solution, Azulay said. “It’s not just sitting on the network layer seeing packets go through a wire. It’s much more sophisticated, with context about the application.”
Bring your own cloud: A strategic differentiator
While traditional SaaS observability vendors charge based on data volume, Groundcover’s “bring your own cloud” architecture allows data to remain within the customer’s environment.
This strategy ensures that a customer’s data remains within its own infrastructure, addressing growing concerns about data privacy and reducing unnecessary data transfer costs. The approach eliminates the need to limit monitoring based on cost concerns, enabling customers to collect up to 10 times more observability data without inflating their budgets, Groundcover asserts.
“We intuitively went into a position where we cannot store customer data,” Azulay said. “It would eventually create a vicious cycle of charging customers for their own data.”
Dashboards for everyone
Last week at the KubeCon EU 2025 conference, Groundcover announced several significant additions to its platform. Among the additions is support for Kubernetes monitoring as well as advanced log monitoring features.
“For anyone using Kubernetes, which is a lot of our users, eBPF can get really intricate insights into the stack, from disk usage to network usage,” Azulay said.
The company also introduced a new eBPF-based dashboarding system. The dashboard system enables users to create their own dashboard to visualize different infrastructure monitoring metrics.
Funding to accelerate market expansion
With the new funding, Groundcover plans to expand its presence in the U.S. market, where it has already demonstrated success converting customers from established competitors. The company is also exploring expansion into adjacent areas including security monitoring and cloud cost optimization, leveraging the data collection capabilities already built into their platform.
“The past year has proven that we know how to combine product and go-to-market in a way that works very effectively for us,” Azulay said. “Right now, it’s our time to hit the gas, because the most precious thing for Groundcover is probably time – to move as fast as we can, as aggressively as we can, with the technological advantage that we have right now.”
Groundcover at a glance:
- Founded: 2021
- Funding: $60 million
- Investors: Zeev Ventures, Angular Ventures, Heavybit and Jibe Ventures
- Headquarters: Tel Aviv, Israel
- CEO: Shahar Azulay
- What they do: Application performance monitoring and observability.
