Kay Ashaolu

Research Interests

• Formation of software engineering judgment in AI-assisted development
• Accelerated professional growth for early-career software engineers
• Framework development for scalable software engineering education
• Impact of artificial intelligence on junior software engineering roles and career development
• Integration of real-life situations and case-study methodologies into software engineering education

Biography

My work is dedicated to accelerating the growth and effectiveness of aspiring and junior software engineers. My goal is to bridge the gap between traditional academic instruction and the rapidly evolving demands of the software industry, particularly in a landscape increasingly influenced by artificial intelligence and emerging technologies. I strongly emphasize group-based, team-oriented, and community-driven peer learning approaches, which I have found critical for fostering deeper engagement and practical skill development among junior engineers. My current research asks how software engineers become solid as they use AI: how they develop the judgment to direct an AI coding agent rather than simply accept what it produces.

I earned a B.S.E. in Electrical Engineering (Computer Systems & Software concentration) from Princeton University in 2007 and an M.I.M.S. from the UC Berkeley School of Information in 2013. My professional journey spans significant engineering and managerial roles at organizations such as the Centers for Medicare & Medicaid Services, Ancestry, and Pinterest, where I currently serve as an Engineering Manager. Concurrently, I am a Continuing Lecturer at UC Berkeley's School of Information.

Central to my current research are two instruments I have built to study how engineers form judgment when working with AI. The first is Scaffold, a free, open-source Claude Code plugin that runs a plan-first developer workflow: an engineer settles and challenges the plan before an AI agent builds, so the engineer grows their judgment while they ship. The second is a framework of seven building blocks: the recurring task and storage patterns that, together with three external entities, describe nearly every software system. It gives engineers a shared vocabulary for judging what an agent has actually built. Recognizing that artificial intelligence is reshaping the expectations and responsibilities of junior engineers, I use both instruments to study how early-career engineers can build real software and grow their judgment on the same timeline, rather than trading one for the other.

Additionally, my scholarly work explores methods for integrating real-life situations and complex problems into the classroom through a structured case-study approach inspired by MBA education models. This practice-oriented method aims to better prepare junior engineers for the dynamic challenges they encounter in professional environments.