Elastic Universal Profiling™ is an emerging technology in observability that allows continuous profiling of code in production environments across various programming languages without the need for service restarts or instrumentation. This system, which can function in both containerized and non-containerized settings, uses eBPF to capture essential performance data with minimal resource consumption, ensuring a low overhead of less than 1% CPU and 250MB of RAM. Unlike traditional profiling tools that often require per-app instrumentation and result in high overheads, Elastic Universal Profiling offers a comprehensive view of the entire system, allowing users to efficiently identify optimization opportunities across large fleets of ARM or x86 nodes. With powerful filtering capabilities and a user-friendly interface, it enables detailed analysis through features like flamegraphs, supporting total observability and effective debugging of performance issues, while also contributing to reduced energy consumption and CO2 footprint.