groundcover Acquires Wand to Build Toward Autonomous Infrastructure

Acquisition brings Kubernetes resource optimization together with groundcover's high-fidelity production context as AI moves from understanding infrastructure to acting on it

SAN FRANCISCO--(BUSINESS WIRE)-- groundcover, the world’s leading bring-your-own-cloud (BYOC), eBPF and OpenTelemetry (OTel)-native observability platform, today announced it has acquired Wand, a Kubernetes resource optimization platform. The acquisition, groundcover's first, moves the company from giving engineers and AI agents the context to understand production toward giving software the ability to safely act on it. Wand's founding team and employees will join groundcover. Terms of the deal were not disclosed.

The acquisition brings together two capabilities groundcover believes will converge as infrastructure becomes more autonomous: the context required to understand production, and the ability to act on it safely.

groundcover captures high-fidelity telemetry across applications and infrastructure using eBPF and OpenTelemetry, retaining that production history inside the customer's own cloud through its BYOC architecture. That data already gives engineers and AI agents the context to investigate issues, understand changes and determine what happened in production. Wand adds a different capability: software that uses production behavior to continuously make and execute decisions about live Kubernetes infrastructure.

"Observability has become much more than a tool engineers use to understand what went wrong," said Shahar Azulay, CEO and co-founder of groundcover. "In an AI-native software development cycle, production data becomes context that agents use to build, troubleshoot, fix and operate software. But understanding is only part of the problem. For infrastructure to become autonomous, systems also need to act safely on what they know. Wand has spent years building exactly that capability for Kubernetes."

From Workload Sizing to Autonomous Infrastructure

Kubernetes resource management is an early example of that shift, and one where engineering teams are already delegating. Platform teams are handing workload scaling to software, including configurations they spent years tuning by hand, because continuous decisions are not work people can realistically keep up with.

The underlying problem is that Kubernetes asks engineers to set CPU and memory requests in advance, usually before a workload has run under real load. Over-provisioning wastes capacity across every replica and node. Under-provisioning risks throttling and outages when traffic shifts. Even a correct decision decays: a configuration that fit a service in March often describes a service that no longer exists by June.

Existing automation addresses parts of the problem but not the interaction between them. Vertical scaling, horizontal scaling and cluster scaling typically operate independently. Tightening a resource request raises measured utilization, which prompts the horizontal autoscaler to add replicas that may not be needed, which changes how workloads pack onto nodes, which changes what the cluster autoscaler does. Each component can behave correctly by its own logic while the cluster as a whole behaves badly.

Wand coordinates those decisions instead of separating them. Its in-cluster technology continuously analyzes workload behavior and makes vertical, horizontal and cluster-level resource decisions together, evaluating the cluster as a whole rather than one workload at a time. Rather than generating right-sizing recommendations for engineers to implement later, Wand adjusts CPU and memory allocation on live infrastructure as demand changes. It installs as a Helm chart, so teams do not take on another set of thresholds and tuning rules to maintain. Wand optimizes for performance and availability first, with cost efficiency following from running the cluster well and removing unnecessary headroom.

"Right-sizing assumes we know the future," said Shon Lev-Ran, CEO and co-founder of Wand. "Spoiler: we don't. Infrastructure changes continuously. Joining groundcover lets us combine Wand's ability to act on live infrastructure with a much deeper understanding of what is happening across production."

Building the Context Layer for Autonomous Infrastructure

The quality of any autonomous system depends on the context behind its decisions. A workload can look over-provisioned based on the last hour and correctly provisioned based on what happens every Monday morning. A latency increase can look like an infrastructure problem until a deployment five minutes earlier comes into view.

groundcover's architecture captures telemetry across applications and infrastructure without requiring teams to instrument every service first, and retains high-fidelity production data without forcing teams to decide in advance what to sample or discard. That history gives engineers and agents more basis for understanding what normal looks like, what changed, and how one part of production affected another.

groundcover is also extending that context beyond telemetry, connecting production data with the engineering systems where information about code, deployments, incidents and decisions lives. An agent can start with something happening in production and understand not only what changed, but why it changed and what happened afterward.

Wand's engine also runs inside the customer's environment, close to the workloads and telemetry it uses to make decisions. The two companies arrived at similar architectural principles independently: as software takes on more responsibility for live production, both the data informing those decisions and the systems acting on them should stay under the customer's control.

What Comes Next

groundcover will apply Wand's technology in two areas.

First, groundcover will use Wand to make its own BYOC data plane more efficient and resilient. Because groundcover runs inside customer cloud environments, the resources it consumes appear directly on customer cloud bills. Making storage and processing more efficient lowers that cost, which in turn lets customers economically retain more telemetry at full fidelity — giving engineers and agents more context to work with.

Second, groundcover will bring Wand's Kubernetes resource optimization to groundcover customers, extending the platform from observing infrastructure toward operating it.

"We're bringing together the context required to understand production and the technology required to act on it," said Azulay. "You need to see what's happening before you can understand it, and understand it before you can safely act. That's the foundation we're building for autonomous infrastructure."

To learn more about how adding Wand’s capabilities to groundcover changes the game, read our blog post.

About groundcover

groundcover is the full-stack observability platform for engineers & agents, allowing teams to monitor everything they run in their cloud, without compromising on data, control, or cost. It’s deployed inside the customer’s cloud premises and provides complete visibility into applications, infrastructure, and AI workloads without operational overhead. The platform offers unlimited data coverage at a fraction of the cost of legacy observability tools. Founded in 2021, the company is backed by leading global investors, including One Peak, Morgan Stanley Expansion Capital, Zeev Ventures, Angular Ventures, Heavybit and Jibe. Learn more at https://www.groundcover.com.

View source version on businesswire.com: https://www.businesswire.com/news/home/20260924605233/en/

Media Contact
Chris Churilo, VP of Marketing
[email protected]

Source: groundcover

You May Also Be Interested In

• 50-Acre Puget Sound Multifamily Asset Sale Brokered by Institutional Property Advisors
• LTK Reimagines Shopping Discovery, Bringing AI and Real Creator Recommendations Together in One Experience
• KBRA Assigns Preliminary Ratings to UK Logistics 2026-4 DAC