Groundcover
eBPF observability with bring-your-own-cloud — now adding AI-native monitoring for LLM workloads.
About
Groundcover (Tel Aviv, founded 2021 by Shahar Azulay) is a cloud-native observability platform built on eBPF from day one: it collects logs, metrics, and traces straight from the Linux kernel with zero code instrumentation, covering Kubernetes and cloud infrastructure out of the box. Its signature move is bring-your-own-cloud — observability data stays in the customer's own cloud while remaining fully managed, which breaks the legacy pattern of being charged by data volume. After a $35M Series B in April 2025 ($60M total, led by Zeev Ventures), it raised a $100M Series C to build AI-native observability: tracking LLM reasoning paths, token usage, and response quality for agent-based workloads alongside conventional infrastructure.
The honest picture: eBPF-based collection genuinely delivers more data at lower cost than agent-instrumented rivals, and BYOC solves a real privacy/billing grievance with Datadog-style vendors. But eBPF means Linux/Kubernetes-centric — non-K8s estates get less value — and as a younger vendor its ecosystem of integrations is thinner than incumbents. The AI-observability pivot is promising but early.
What it can do
- Zero-instrumentation eBPF collection of logs, metrics, traces
- Bring-your-own-cloud: data stays in your infrastructure
- Kubernetes monitoring, log monitoring, custom eBPF dashboards
- AI-native observability: reasoning paths, token usage, response quality
- Usage-based pricing decoupled from data volume
Setup
Deploy the Groundcover Helm chart or operator into your Kubernetes cluster; eBPF sensors start collecting immediately with no code changes. Connect your cloud account for the BYOC data plane.
Prerequisites
A Kubernetes cluster (Linux nodes with eBPF support); a Groundcover account.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 58; Trending score 74/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.