Abstract
The field of Artificial Intelligence (AI) has rapidly transformed from a buzzword technology to a fundamental aspect of numerous industrial software applications. However, this quick transition has not allowed for the development of robust best practices for designing and implementing processes related to data engineering, machine learning (ML)-based model training, deployment, monitoring, and maintenance. Additionally, the shift from academic experiments to industrial applications has resulted in collaborative development between AI engineers and software engineers who have reduced expertise in established practices for creating highly scalable and easily maintainable processes related to ML models. In this paper, we propose a series of good practices that have been developed as the result of the collaboration between our team of academic researchers in AI and a company specializing in industrial software engineering. We outline the challenges faced and describe the solutions we designed and implemented by surveying the literature and deriving new practices based on our experience.
Citare
@Inproceedings{Moroz-Dubenco2023TowardsGP,
author = {Cristiana Moroz-Dubenco and B. Mursa and Mátyás Kuti-Kreszács},
booktitle = {International Conference on Software and Data Technologies},
title = {Towards Good Practices for Collaborative Development of ML-Based Systems},
year = {2023}
}