talks

GenAI is rapidly reshaping software engineering, advancing capabilities in code generation, translation, testing, and issue analysis. However, current evaluation practices remain fragmented, inconsistent, and often irreproducible, making it difficult to assess genuine progress. This talk explores the challenges of systematically and transparently benchmarking GenAI for software engineering. We present a unified framework that integrates key components (metrics, workloads, prompting strategies, and experimental procedures) to enable rigorous and comparable assessments across diverse tasks.

Serving as the Doctoral Symposium Keynote speaker at ISSREW 2025, this talk shares essential lessons, methodologies, and career strategies for PhD candidates navigating the field of software reliability. Professor Vieira outlines critical pathways for successful research, emphasizing experimental design, collaborative networks, and addressing modern reliability challenges in AI-driven systems.

Large Language Models (LLMs) are transforming software engineering, offering new possibilities for developing secure and trustworthy software. This keynote explores the integration of LLMs into software development workflows, particularly their role in code generation. Supported by empirical evidence, we discuss the capabilities of LLMs in vulnerability detection and mitigation, and delve into the importance of assessing the trustworthiness of code, including the role of LLMs in verifying code quality and adherence to best practices.

Vulnerability detection is crucial to securing modern complex software systems. This keynote reviews recent paradigm shifts and engineering advances in vulnerability detection, ranging from static and dynamic analysis to the combination of multiple security tooling alerts. Supported by empirical research, Vieira discusses how combining analysis techniques and leveraging machine learning can significantly reduce false-positive rates and help developers focus on critical fixes.

Benchmarking has successfully driven performance improvements in systems for decades, but applying it to dependability and security remains exceptionally challenging. This keynote critically examines the state of the art in security and dependability benchmarking. Vieira explores the essential prerequisites, grand challenges, metrics, and tools required to make security assessments repeatable, reproducible, and trustworthy across industry and academia.