Mr. Aditya Gupta

Causal Debugging and Intelligent Tracing: Rethinking Observability for Modern Systems

Abstract:

In today’s multi-service architectures, developers are drowning in telemetry data but still struggling to answer the most basic question: why did something fail? This talk introduces a practical framework for causal debugging using structured tracing, span enrichment, and failure signal inference. We will explore how to think beyond dashboards and alerts, and instead build intelligent tracing systems that reflect causal chains and surface latent dependencies in runtime workflows. The session presents low-overhead design patterns such as metadata propagation, minimal span transforms, and trace-query models that help correlate infrastructure issues with real user or business impact. This is not just a theory talk. The techniques shared are field-tested and grounded in trace observability research. Whether you are maintaining distributed systems, working on high-throughput APIs, or scaling infrastructure visibility, this talk will help you design tracing with intent, not just afterthought.