Krishna Ronanki

PhD Student - Software Engineering | Trustworthy AI | AI Adoption | AI Engineering


Research Abstract

Large language models (LLMs) are rapidly reshaping software engineering (SE), evolving from chatbots used ad hoc by developers to integrated components of SE workflows and, most recently, to multi-agent systems that perform SE tasks with growing autonomy. Industrial adoption of LLMs keeps growing while the understanding needed to support it is still lacking in terms of methods for assessing LLM output reliability, LLM-specific practices and actionable guidance in adoption models. Meanwhile, trustworthy AI principles, now reinforced by the EU AI Act, remain hard to operationalise. My research aims to understand the evolution of LLM adoption in SE and to enable its trustworthy adoption by focusing on three major aspects: evaluating LLM capability in SE tasks, understanding and guiding adoption in industrial practice, and enabling trustworthy adoption.

Key Areas of Focus

1. Inferring evolving insights regarding the applicability of LLMs in software development.2. Identifying the key barriers and enablers influencing the trustworthy adoption of LLMs.3. Synthesising artefacts and mechanisms that facilitate the trustworthy adoption of LLMs to enhance AI-augmented software development processes.