Software Engineering for AI Systems

The scope of the CAIN conference is Software Engineering for AI Systems. While there are many well-established venues in the fields of artificial intelligence (AI) and machine learning (ML), CAIN is unique in that it takes a systems and life cycle perspective on AI engineering, and we want to have a space for deep discussions, proposals, and solutions in this area. Therefore, all submissions must have a software engineering and systems focus.

We consider three types of submissions to be within the scope of CAIN:

• AI Systems Submissions: These are submissions that fully fall within the scope of CAIN. In these submissions, software engineering methods and practices are applied/expanded/reconsidered for the engineering AI-enabled systems, or novel approaches are introduced to promote system qualities of AI-enabled systems (e.g., bias, fairness, explainability).
• AI Component Submissions: While not fully AI Systems submissions, these submissions apply software engineering methods and practices to AI components (e.g., software testing for ML components). Submissions focusing exclusively on the AI component, with no systems or software engineering focus, will be desk rejected (e.g., optimization of hyper-parameter selection).
• AI4SE Submissions: AI4SE is a very large field with its own set of venues. We will only accept AI4SE submissions in which AI supports developers in using software engineering methods and practices for AI systems or components (e.g., test case generation based on LLMs).