Intelligent Software Engineering in the era of Agentic AI

About the Project

Given the tremendous development of AI in recent years, innovations like the Agentic AI that is based on Large Language Model (LLM), e.g., ChartGPT, have significantly impacted our daily lives. Indeed, Agentic AI does not only change the understanding of what we think software can do but also creates a paradigm shift on how we should engineer software systems. This project is thus motivated under such a context, aiming to explore cutting-edge methods that underpin the synergy between Agentic AI and Software Engineering (SE). The concrete project topic is rather open but will sit in one of the two themes: Agentic AI for SE or SE for Agentic AI.

In Agentic AI for SE, the project will investigate how Agentic AI can be better tailored and turned into a more specialized variant, hence better solving the software engineering tasks at hand. There is no constraint on the software engineering task to solve, which can include, but is not limited to, software defect prediction, bug report analysis, configuration performance learning/tuning, and software testing. Under the SE for Agentic AI, this project will look at novel ways that can properly evaluate, test, and verify the different properties of Agentic AI-assisted tools, including their robustness, fairness, and security. In essence, for intelligent software engineering in the era of Agentic AI, both themes in the project will work on a creative combination of software engineering knowledge with the characteristics of Agentic AI, and such a unique combination will inevitably lead to both challenges and opportunities, making this research as a fruitful direction for a PhD topic.

The deliverable of the project can be of diverse forms, including methodology, algorithms, or the actual prototype of the tool, depending on the identified nature of the research. You will work in a group of 8 PhD students under close supervision with the chance to work with several industrial collaborators of the group. We offer generous support and a friendly environment. The ideal candidate should have knowledge of AI and/or SE, but it is not essential. Interested applicants are strongly advised to contact Dr. Chen to discuss more details (t.chen@bham.ac.uk) and more information can be found on IDEAS lab (https://ideas-labo.github.io/).

When contacting, please send over three information: (1) your CV; (2) a research proposal; and (3) why you would like to join the group.

Eligibility:

First undergraduate degree and/or postgraduate degree with Distinction (or an international equivalent). The applicant should have a degree on Computer Science, Artificial Intelligence, and/or Software Engineering. Applicants with peer-reviewed publications are strongly encouraged to apply. We also welcome applicants with a self-funded basis.

Application can be made here: https://www.birmingham.ac.uk/schools/computer-science/postgraduate-research/applying-for-phd-in-computer-science