Pragun Saxena enhanced TCP transport observability and performance across the Intel-tensorflow/tensorflow, Intel-tensorflow/xla, and grpc/grpc repositories by expanding TCP metrics, improving timestamp formatting, and centralizing tracing controls for third-party integrations. Working in C++ with a focus on backend and network programming, Pragun introduced Xprof profiling instrumentation and refined event coverage to support detailed transport profiling. Tests were updated to reflect new metrics, and feature flags were implemented to guard statistic collection, reducing noise and improving reliability. These changes enabled faster mean time to resolution, better SLA visibility, and data-driven optimization of the transport stack in cloud-native environments.