Platform teams managing instrumentation across hundreds of distributed services face a common challenge: achieving observability coverage without creating maintenance nightmares.

Dealing with polyglot environments, inconsistent spans, and the high overhead of manual instrumentation often results in fragmented telemetry and "0-to-1" implementation gaps.

This talk moves past the basics to compare the three pillars of modern instrumentation: SDK-based auto-instrumentation for rapid "0-to-1" wins, the OpenTelemetry Kubernetes Operator for platform-wide injection, and eBPF for visibility in environments where code changes are impractical.

This session covers practical guidance on auto-instrumentation and the transition to high-value custom instrumentation, ensuring the telemetry metadata required for rapid troubleshooting in production is captured.

Attendees will leave with deployment patterns & decision frameworks for balancing instrumentation coverage against operational complexity.