Lokeswari Vejju

Minimum-Exposure Observability: Balancing Diagnostic Utility, Security, and Privacy

Abstract:

Modern observability systems rely on logs, metrics, and distributed traces to diagnose failures and maintain reliable software systems. In security-sensitive environments, however, excessive telemetry can increase privacy, security, and governance risks. This talk explores the concept of minimum-exposure observability and examines how telemetry minimization can affect diagnostic utility across different operational tasks. Drawing on recent research, the talk will discuss practical approaches such as pseudonymization, selective data reduction, sampling, aggregation, and policy-based telemetry controls, along with guidance for balancing effective diagnosis with reduced data exposure.

Profile:

Lokeswari Vejju is an Independent Researcher and Software Development Engineer II with more than six years of professional experience in secure cloud infrastructure, cryptographic systems, distributed systems, observability, performance engineering, and financial technology. Her professional experience spans Amazon Web Services, Goldman Sachs, and Citi. Her recent research focuses on secure and privacy-conscious observability for distributed systems. She holds an M.S. in Computer Science from Arizona State University and a B.Tech. in Information Technology from NIT Allahabad.
 

Download profile as PDF: Click Here