Back

ⓘ MLOps is a practice for collaboration and communication between data scientists and operations professionals to help manage production ML lifecycle. Similar to ..




                                     

ⓘ MLOps

MLOps is a practice for collaboration and communication between data scientists and operations professionals to help manage production ML lifecycle. Similar to the DevOps or DataOps approaches, MLOps looks to increase automation and improve the quality of production ML while also focusing on business and regulatory requirements. While MLOps also started as a set of best practices, it is slowly evolving into an independent approach to ML lifecycle management. MLOps applies to the entire lifecycle - from integrating with model generation, orchestration, and deployment, to health, diagnostics, governance, and business metrics.

                                     

1. History

The challenges of the ongoing use of machine learning in applications were highlighted in a 2015 paper titled, Hidden Technical Debt in Machine Learning Systems.

The predicted growth in machine learning includes an estimated doubling of ML pilots and implementations from 2017 to 2018, and again from 2018 to 2020. Spending on machine learning is estimated to reach $57.6 billion by 2021, a compound annual growth rate CAGR of 50.1%.

Reports show a majority up to 88% of corporate AI initiatives are struggling to move beyond test stages. However, those organizations that actually put AI and machine learning into production saw a 3-15% profit margin increases.

In 2018, MLOps and approaches to it began to gain traction among AI/ML experts, companies, and technology journalists as a solution that can address the complexity and growth of machine learning in businesses.

                                     

2. Architecture

There are a number of barriers that prevent organizations from successfully implementing ML across the enterprise, including difficulties with:

  • Scalability
  • Diagnostics
  • Collaboration
  • Deployment and automation
  • Governance and regulatory compliance
  • Reproducibility of models and predictions
  • Business uses

A standard practice, such as MLOps, takes into account each of the aforementioned areas, which can help enterprises optimize workflows and avoid issues during implementation.

A common architecture of an MLOps system would include data science platforms where models are constructed and the analytical engines were computations are performed, with the MLOps tool orchestrating the movement of machine learning models, data and outcomes between the systems.