May 13
Loading video player...
Observability is often treated as an operations afterthought. But by the time production issues hit, missing traces and half-instrumented code leave ops teams flying blind. We learned this the hard way when critical API endpoints had no telemetry, turning incidents into guesswork. In this talk, we show why observability must start in development and how platform and operations teams can drive that change. We will walk through several OpenTelemetry features and explain why developers should use them during development. These features include child spans, database query instrumentation, outgoing HTTP traces, trace-log correlation, and error propagation. We explain how early adoption reduces bugs, eliminates production blind spots, and speeds up incident response. You will learn concrete techniques to equip developers with the right telemetry practices, set environment-specific instrumentation strategies, and build a stronger Dev/Ops collaboration. This talk is for platform engineers, SREs, DevOps leaders, and developers who want to improve reliability and eliminate guesswork in production by baking observability in from day one.