Dash0 Acquires Polar Signals for Continuous Profiling and GPU Visibility

Observability startup Dash0 has acquired Berlin-based continuous profiling specialist Polar Signals, adding continuous profiling capabilities to its SignalStore data platform. Continuous profiling, which has been part of the developer toolbox for years, provides an ongoing view of CPU usage, memory allocation and resource consumption within running applications.

Polar Signals offers a unique feature: continuous profiling of Nvidia CUDA workloads in production, tracking AI training and inference workloads at an individual GPU kernel level for monitoring resource consumption and pinpointing performance issues. Dash0 will integrate Polar Signals' profiling data into SignalStore, feeding it into Agent0, its AI agent for production operations.

Agent0 can now correlate telemetry with source code and prepare pull requests for fixes, with profiling adding another source of runtime evidence for developers to consider. Dash0 will also allow external coding agents access to the profiling data through its MCP server, providing runtime context for suggesting changes. The acquisition was the result of a conversation at a dinner in Berlin between Dash0 CEO Mirko Novakovic and Polar Signals founder Frederic Branczyk, who recognized their companies were working towards the same goal with different approaches.

This overlap suggests the deal extends beyond continuous profiling, as Dash0 plans to replace ClickHouse beneath SignalStore with Polar Signals' Great Lakes storage engine, handling high-cardinality data generated by profiling. Polar Signals' open source continuous profiling project, Parca, will remain open source and continue to be maintained.

The acquisition, which did not disclose financial terms, further complements Dash0's OpenTelemetry-based architecture, with OpenTelemetry Profiles entering public alpha in March, extending the OpenTelemetry data model to include profiles alongside traces, metrics and logs.