Pixie is an open source observability tool for Kubernetes applications that uses eBPF to automatically collect telemetry data without the need for manual instrumentation. Developers can access metrics, events, traces, and logs in seconds without changing application code, using dynamic eBPF probes and ingestors.

The platform runs entirely inside Kubernetes clusters, collecting, storing, and querying all telemetry data locally without moving data outside the cluster. This in-cluster edge compute model means Pixie uses less than 5% of cluster CPU and in most cases less than 2%, enabling deep visibility without significant performance overhead.

Pixie supports a flexible Pythonic query language called PxL, which can be used across the web UI, CLI, and client APIs to run community, team, or custom scripts for debugging and analysis. Developers can investigate service maps, cluster resources, application traffic, pod states, flame graphs, and individual full-body application requests through a unified interface.

The platform supports exporting data in the OpenTelemetry format, enabling integration with external tools for long-term data retention and alerting. Pixie is a Cloud Native Computing Foundation sandbox project, open sourced under the Apache 2.0 license, with a hosted managed version available at no cost.