Ten Things to Know About ModelOps

Chapter 1. ModelOps

The operationalization of AI and machine learning (AI/ML) models has become a critical topic for enterprises to address: the significant advances achieved in recent years with respect to algorithms and technologies around data science, AI, ML, and advanced analytics have enabled entirely new use cases. AI/ML can optimize, streamline, and significantly improve critical processes for virtually all organizations across all domains from manufacturing, financial services, marketing, customer relations management, healthcare, logistics, and so on. But with the wider adoption of AI/ML also come greater cost, risk, regulation, and demands on organizational processes, new tools, tasks, and teams.

Today, many organizations rely on dozens or hundreds of AI/ML models. Almost all organizations are in various stages of planning or implementing systems and processes to embrace AI/ML. The effective management of those models, their life cycles from development through agile deployment and monitoring, their governance, and the effective implementation to “improve solutions” have become a critical concern. In summary, that is the purpose of ModelOps applications, methods, and solutions.

The authors of this book bring decades of combined experience and real-world learning on how to overcome the challenges of implementing data science and analytics applications in practice, for real return on investment (ROI), at scale, and consistent with best practices, regulatory constraints, ...