IDI resident research software engineer holds agentic coding workshops: Command Prompt
Over the past few months, the Institute’s resident research software engineer, Evan Morgan, has been organising meetups and workshops for staff and students across the university to discuss and learn about AI-assisted coding. These events were motivated by the fact that software development and coding practices have experienced some of the fastest and most profound impacts of advances in generative AI. Software engineers, many of whom spent years gaining expertise in how to write code, are now finding that AI agents can do most of this work for them. Conversely, those with no experience of coding are suddenly able to prompt their way to functional software in the space of hours.
Evan established Command Prompt as a monthly forum for the university community to discuss these impacts, share skills and resources, and provide guidance on the latest tools and best practices. The meetup drew over 200 attendees across the first three events, where presenters and attendees discussed workflows for AI-assisted coding, security best practices, and the need for clearer policies and guidance around the use of agentic coding tools.
As a ‘Researcher in Residence’ with the university’s Digital Research Services, Evan also received funding to facilitate the delivery of two day-long workshops on Agentic Coding. The workshops were attended by over 100 staff, including those with little software development experience. Attendees spent the morning learning about AI coding tools and workflows, before putting their new skills to test in an afternoon hackathon.
These events have highlighted generative AI’s transformative potential for enabling innovation across a spectrum from experimental research code, through to production and open-source software solutions. At the same time, there is a clear need (and appetite) for sufficient training, resources, and guardrails to ensure these rapidly advancing tools are employed responsibly.
