Mr. Aditya Gupta

Trace-First Observability: Blueprints for Modern Systems

Abstract:

AI features rarely fail quietly, they fail expensively. This talk teaches a trace-first approach to observability so you can see latency spikes, model hot paths, cache misses, and downstream timeouts before customers do. We will instrument a service with OpenTelemetry from SDK auto instrumentation to Collector pipelines, then export to AWS X-Ray using ADOT and to Google Cloud Trace in Google Cloud Operations. We will containerize the example with Docker and show how traces, metrics, and logs work together to tie model versions to error budgets and to catch regressions. An LLM assistant will sit on top of spans to speed incident triage with privacy and PII guardrails. For security, we will run a fast STRIDE pass, apply image signing, registry scanning, policy checks in CI, and detect anomalous egress using trace signals. You will leave with copy pastable blueprints, including OpenTelemetry configurations, Dockerfile patterns, and dashboards that make systems observable, explainable, and secure by design.

You will learn

• Practical OpenTelemetry wiring to AWS X-Ray and Google Cloud Trace
• How to use traces as the primary signal for debugging and performance tuning
• A repeatable threat model and CI policy workflow for safer releases

Profile:

Aditya Gupta is a Senior Member of IEEE and distinguished cloud observability engineer at the world's largest cloud service provider, where he leads critical infrastructure processing over 1 million transactions per second for global enterprises and government agencies. As a pioneering researcher in machine learning and distributed systems, he has authored 15+ peer-reviewed publications with multiple citations across premier IEEE, Springer, and Elsevier venues, establishing foundational methodologies in XGBoost validation that are cited globally across healthcare, environmental science, and educational technology. Over the past decade, he has conducted numerous peer reviews for top-tier journals and conferences, earning recognition as a top-percentile reviewer in computer science. Mr. Gupta serves as Session Chair at multiple international IEEE and Springer conferences and is an invited speaker at premier venues across North America, Europe, and Asia. He holds fellowships in prestigious organizations, while serving as a judge for U.S. Department of State fellowship programs and international innovation competitions. His technical expertise spans cloud observability, distributed systems, machine learning, and cybersecurity, with thought leadership articles reaching millions of readers across premier developer platforms. Mr. Gupta represents the intersection of academic rigor and industry innovation, with measurable impact on both global research communities and critical infrastructure systems serving millions worldwide.