Boost machine learning development operations value with innovation management integration

Despite the increasing adoption of machine learning across various sectors, many organizations, especially Small and Medium Enterprises (SMEs), struggle to scale machine learning initiatives into production. This paper proposes integrating Innovation Management into Machine Learning Operations (MLOps) to enhance organizations' ability to capture the true business value of machine learning. MLOps, a discipline comprising tools and practices, addresses the challenges of deploying machine learning products in real-world scenarios. Using a qualitative methodology, we analyze interviews, academic sources, and books to construct a best practice matrix aligned with both MLOps principles and innovation management. Findings reveal that innovation-oriented practices significantly enhance the value extraction, adaptability, and performance of machine learning deployments.