research

ICET-lab studies two connected questions: how to measure, predict, and improve the performance of software systems, and how developers and organisations actually work with LLMs and coding agents. Increasingly these are the same question, since code is now produced at machine speed by systems whose efficiency is rarely checked.

Many of our studies have an empirical slant (e.g., we crawl open source repositories, send out surveys, or conduct interviews), but we also frequently conduct experimental research (e.g., controlled lab experiments, field studies, or performance evaluations). We are also keen on building proof-of-concepts and tools. We are commited to open and reproducible science, as well as to transferring our results to industry and society as a whole. As part of this, we often speak about our work at academic and industrial conferences.

Virtually all our papers are available freely via the green open access model. If you cannot find a specific paper you are always free to email the authors, we will happily provide a (text-identical) preprint version.

ICET-lab is currently mostly funded through grants from the Swedish Research Council (VR) and the Wallenberg AI, Autonomous Systems, and Software Programme (WASP). Historically, we have also received significant financial support from Sweden’s innovation agency Vinnova, the Swiss National Science Foundation (SNF), and the ICT Area of Advance at Chalmers, as well as the EU Horizon 2020 research programme.