🤖About Me
Roham Koohestani is a Software Engineer and AI4SE researcher based in The Hague, Netherlands. He studies Computer Science & Engineering at TU Delft and works at the intersection of large language models and developer tooling. Roham is driven by a core question: how can we build AI tools that truly support developers rather than distract them?
His work spans theory, experiment, and system-building; from modeling developer behavior and integrating AI into IDE workflows to empirically measuring trust, calibration, and usefulness in human–AI collaboration. He has collaborated across academia and industry (TU Delft, JetBrains Research, and others) to translate research insights into developer-facing tools and publications.
Looking ahead, Roham aims to pursue a PhD to advance the scientific foundations and real-world impact of AI-augmented software development while maintaining a strong engineering mindset.
Education
BSc Computer Science & Engineering
TU Delft
Minor in Mathematics
University of Amsterdam
Natuur & Techniek
Het College Weert
Research Interests
I specialize in Artificial Intelligence for Software Engineering (AI4SE): applying ML and large language models to improve how software is written, reviewed, maintained, and evolved. My work spans theory, experiment, and system-building; including, but not limited to, modelling developer behaviors and integrating AI into IDE workflows to empirically measuring trust, calibration, and usefulness in human–AI collaboration.
I collaborate across academia and industry (TU Delft, JetBrains Research, and others) to translate research insights into developer‑facing tools and publications. Looking ahead, I aim to pursue a PhD to advance calibrated, trustworthy, and impactful AI‑augmented software development. If you’re interested in collaborating on LLMs‑for‑Code, developer–AI interaction, or agentic systems, feel free to get in touch.