groundcover raises $100 million to challenge Datadog in AI observability
The AI observability startup groundcover has secured $100 million in new funding to challenge established players like Datadog by offering a host-based pricing model tailored to the massive data demands of artificial intelligence workloads.
groundcover has raised $100 million in a round led by One Peak, bringing its total capital to $160 million. The four-year-old startup reports more than 250 paying customers and says it has tripled its annual recurring revenue over the past year.
The funding highlights a broader shift in the enterprise software market as artificial intelligence generates unprecedented volumes of operational data. groundcover is using this capital to challenge established observability giants like Datadog, Dynatrace, New Relic, Splunk and Grafana, which have long dominated the sector with billions in annual revenue.
Traditional observability platforms typically charge customers based on the volume of data ingested. However, autonomous AI agents and complex distributed systems produce massive amounts of telemetry, including prompt execution, model latency and token consumption.
This pricing model forces engineers to sample or limit data collection to control costs. "We've seen telemetry exploding," groundcover co-founder and CEO Shahar Azulay said. "Users are frustrated by not getting all the value from Datadog and similar platforms."
To counter this, groundcover utilizes a bring-your-own-cloud architecture where the data plane remains inside the customer's own AWS, Azure or Google Cloud environment. The company charges based on the number of monitored hosts rather than data volume, offering a fully self-hosted deployment option as well.
"We don't price by data volume," Azulay said. "We price by the size of the infrastructure." This model aims to provide predictable billing for enterprise teams struggling with fluctuating observability costs.
Technical differentiation
The platform relies heavily on eBPF, a Linux kernel technology that observes network traffic and system behavior without requiring manual code instrumentation. This approach significantly reduces deployment complexity and shortens implementation times for organizations running dense Kubernetes clusters.
"Our sensor allows us to observe systems very deeply from infrastructure to application to AI workloads without developers needing to instrument code," Azulay said. The company combines this automatic collection with OpenTelemetry compatibility and managed control planes.
groundcover also envisions a future where observability platforms serve autonomous AI agents rather than just human operators investigating production incidents. Whether this architectural shift will successfully displace incumbent vendors remains an open question, but the rapid revenue growth suggests strong market demand for alternative models.
