2024

Ahmad Tazeem; Adnan Mohd; Rafi Saima; Akbar Muhammad Azeem; Anwar Ayesha

In recent years, the significant increase in enterprise data availability and the progress in Artificial Intelligence (AI) have enabled organizations to address real-world issues through Machine Learning (ML). In this regard, machine learning operations (MLOps) have been proven to be a beneficial strategy for evolving ML models from theoretical ideas to practical solutions of business sector issues. With the knowledge of MLOps being vast and scattered, this research work focuses on the application of MLOps methodologies in sophisticated operational technologies, prioritizing the enhancement of security measures. This research work also discusses the specific challenges in securing ML implementations in such settings and underscores the importance of robust MLOps strategies in ensuring effective security protocols. Moreover, it explains current practices and identified improvement areas, highlighting the importance of MLOps in overcoming obstacles and maximizing the value of ML in operational technology contexts.