Software Engineering Research

Software Engineering
Research

Driving Innovation in Software Engineering

We are a team of researchers and engineers bringing innovative technology and scientific research to JetBrains products.

Research directions

Our research directions are defined by our vision of the future of software engineering tooling, roadmaps of JetBrains products, and cutting-edge problems in the academic research community.

Each card links to the team's webpage with more information.

AI in software engineering

We use machine learning techniques and agentic approaches to help developers and enhance software development processes. We aim to improve the adoption of AI-enabled coding tools by turning the latest academic advances into practical applications.

Human-AI experience

We study the impact of AI on software developers and their work, design new modes of human-AI interaction, and search for ways to evaluate the utility of AI tools in software engineering.

Software testing research

We are working to make software testing easier and more powerful by designing new techniques and tools.

Education research

We design and develop innovative features for learning tools.

Former research labs

Select projects

Our research spans different topics, with both scientific and technical approaches.

Here are a few of our ongoing projects.

Federated compute

Applying machine learning effectively to software development demands insights from extensive, real-world codebases. However, conventional centralized approaches frequently put data privacy at risk. To address this, we are investigating federated learning methods that enable efficient, privacy-sensitive solutions.

Find out more

Proactive and smart AI chat

While early studies on proactive AI agents within their IDEs show the potential for proactive coding assistants, these studies have mostly used simple models and haven't addressed real-world challenges like cost and privacy. This project aims to develop a new prototype that intelligently delivers proactive suggestions, enhancing the user experience and providing insights into building trigger models that mimic human cognition.

IdeGYM

IdeGYM is a framework that quickly creates disposable coding environments for AI agents or other machine learning tools. It can launch thousands of ready-to-use workspaces with IntelliJ IDEA in seconds, enabling efficient training, testing, and scaling of AI projects.

AI debugging

The main goal of the project is to develop a plugin that actively guides students through the debugging process while they solve coding problems. This plugin will offer specific advice what breakpoints to set and where, which variable values to monitor, and provide other helpful debugging tips to enhance their learning experience.

Find out more

Academic collaborations

Many of our research projects are run jointly with our academic partners and collaborators.

Community contributions

We contribute to the scientific community in a number of ways.

Publications

We aim to ensure that the community can build on our work by publishing scientific papers, based both on our in-house research and collaborations. Our Google Scholar page

Community service

We take part in scientific community service by reviewing papers and organizing academic events in leading venues.

Sponsorships

We support the community by sponsoring numerous academic events.

Internships

We provide a learning environment for early-career researchers from all over the world by hosting research internships.

If you are interested in collaborating with us, please reach out!

Please refer to individual team pages for more details on open collaboration opportunities.