Recommendations for Developing Machine Learning Operations (MLOps) 


Algorithms that automatically learn and improve decision-making without additional training use a process known as machine learning (ML), a field that represents a rapidly developing branch of artificial intelligence (AI). By 2022, the ML market is expected to skyrocket to USD $8.81 billion (up from about USD $1 billion five years ago). 

Powered by data, ML is growing in popularity due to its ability to help businesses operate faster and more efficiently, as well as make more informed decisions. Marketers use it to target ad campaigns at specific audiences. Social media companies leverage it to spot and remove malicious content from their platforms. Search engines like Google deploy it to boost search functions. Its potential benefits for companies developing futuristic technology, from robots to autonomous vehicles, is enormous.

How to Create A Governance Framework for Machine Learning

Many organizations today are focused on implementing ML, but few have adopted appropriate best practices and policies to make the technology work for them operationally.

Creating governance around machine learning — often referred to as “ML operations,” or simply “MLOps” — is important for any organization that wants to turn its machine learning aspirations into reality. During a recent interview with Forbes, Jamal Robinson, who heads AI/ML Business Development at Amazon Web Services, said that “50% of ML projects never see the light of day.” As ML continues to evolve, many businesses lack “governance structures required to make ML explainable, repeatable and production ready,” he said.

Part of the problem is that organizations tend to be overly fixated on the technical aspects of machine learning. Instead, they should be determining the operational ones, such as project management, cost controls necessary for ML experimentation, the adoption of a strong platform that appeals widely to different stakeholders (like data scientists and business analysts), setting up an explainable ML model development process, and implementing ML model and data governance frameworks. 

“In addition to data quality, data governance requires an understanding and development of best practices for metadata management, storage, integration, interoperability and securing access to data, amongst other things,” Robinson said. He recommends organizations that want to develop a sound MLOps consider the following:

  • Review the Data Management Body of Knowledge (DMBoK) and related materials to better understand what a comprehensive data governance framework should look like. 
  • While there is presently little offered in terms of specific industry-wide standards on model governance, some related materials organizational leaders should reference include Ernst & Young’s Model Risk Management for AI & ML, Singapore’s Model AI Governance Framework, and Amazon Web Services’ Machine Learning Best Practices in Financial Services.
  • Before a project launches, develop a method for categorizing ML projects related to your organization’s ability to execute an anticipated return on investment. To begin, leaders should put ML projects into one of three categories based on their desired business goals: efficiency improvements, automating decision making, and innovation. 
  • Lastly, he said businesses with less mature ML should begin with “efficiency improvements or automating decisions,” areas where most organizations already invest and are often able to succeed.

ML holds much potential for organizations that want to improve their products and operations. By implementing appropriate MLOps around the technology, organizations will be much better able to adapt to the many changes it will bring.

Understand Machine Learning

By providing AI with the ability to learn from its experiences without needing explicit programming, machine learning is important to developing the technology. Covering machine learning models, algorithms, and platforms, Machine Learning: Predictive Analysis for Business Decisions, is a five-course program from IEEE.

Connect with an IEEE Content Specialist today to learn more about this program and how to get access to it for your organization.

Interested in the program for yourself? Visit the IEEE Learning Network.


Drenik, Gary. (17 July 2021). Importance Of Data, Governance And MLOps When Using Machine Learning To Drive Successful Business Outcomes. Forbes.

Lazzaro, Sage. (21 June 2021). Machine learning’s rise, applications, and challenges. Venture Beat. 

, ,


  1. Four Ways to Build MLOps that Avoid “Data Drift” in Machine Learning - IEEE Innovation at Work - August 19, 2021

    […] can prevent data drift by building machine learning operations (MLOps). As we discussed in a previous post, an appropriate MLOps establishes a governance framework around machine learning that helps […]

Leave a Reply