User-Centered In-IDE HAX
With TU Delft: folding emerging LLM practices into the IDE workflow without disturbing the developer.
Explore projectHuman–AI Experience
How AI reshapes the way developers think, collaborate, and get unstuck — what changes, what stays the same, and what to design around.
What we focus on
Design of the interaction
Where AI assistance belongs inside the IDE and across the development lifecycle, so that it fits the workflow a developer already has instead of interrupting it. That means looking past the chat window — proactive and context-aware help, in-place affordances, and low-friction ways to ask for it or receive it.
Impact on developers and teams
How in-IDE AI changes behavior, skill, and long-term tool use — and what it does to collaboration. Interviews, surveys, telemetry, and longitudinal analysis tell us who adopts which features, why others opt out, and how attitudes move as the tools mature.
Quality of interaction and output
Which properties of an AI system actually matter in practice: explanation quality, latency, controllability, predictability, perceived trustworthiness. We build measures of whether a tool is useful in real software engineering, not only of how it scores offline.
Quo Vadis, Code Review?
By Michael Dorner
The people who succeed with AI tooling don't generate more — they verify faster.
Projects
Preference Elicitation for AI Agents
With TU Delft and UC Davis: which model characteristics developers actually value, and how that shifts by task.
Explore project
Code Review for AI-Generated Code
With Lund University: review workflows built for appropriate trust in AI-generated changes, not blanket acceptance.
Explore project
Reasoning Trace Continuation
With UC Irvine: how people and LLM agents hand a problem back and forth — and what the handover does to the answer.
Explore project
Selected publications
Apr 2026
StudyHuman–AI Experience
Developer Needs and Feasible Features for AI Assistants in IDEs
Asks developers what they actually want from in-IDE assistance, then sorts those wants by what is feasible to build.
Jan 2026
ReviewHuman–AI Experience
Human-AI Experience in Integrated Development Environments: A Systematic Literature Review
Maps what is known about human–AI interaction inside the IDE, and where the field's evidence actually stops.
Jun 2025
StudyHuman–AI Experience
AI in Software Engineering: Perceived Roles and Their Impact on Adoption
Developers cast AI tools either as an inanimate instrument or as a human-like teammate — and the more roles they assign one, the more useful they find it.
Other directions
Research that ships
Applied Research
Prototypes, studies, and instruments that bridge research findings into the products millions of developers use every day.
Beyond today's models
Machine Learning
Core ML research that powers verification, generation, and reasoning in our tools — papers, benchmarks, and the teams behind them.