MLops is the hot new cloud computing career path

ClearML, an open source MLops platform, announced its new research report: “MLOps in 2023: What does the future hold?” This study surveyed 200 machine learning decision makers in the United States, examining key trends in machine learning and MLops (machine learning operations).

Potential vendor self-service bias aside for now, the ClearML study found that MLops now enjoys wide-scale adoption within enterprises; 85% of respondents said they have a budget dedicated to MLops in 2022. And 14% said they don’t have budgets in place but expect to have them in 2023. So companies are going MLops now or soon.

In case you haven’t noticed, operations seems to be the new focus of cloud computing work. We have cloudops (cloud operations), finops (financial operations), devops (development and operations), and secops (security operations). You can see the trend.

This is for a good reason. The creation and implementation of cloud solutions or the migration of existing solutions to the cloud are necessary tasks. Normally, they are one and done. The focus then shifts to operations to keep the value of that work coming back to the business. As many companies have discovered in recent years, simply throwing things at a public cloud provider and hoping for the best returns no value. Neglecting operations, all operations, leads to huge cost overruns and little return on investment.

MLops is a critical component of the machine learning lifecycle, enabling organizations to manage and operate machine learning models in production. MLops processes ensure that models are consistently and efficiently deployed, monitored, and updated, allowing organizations to reap the full benefits of machine learning. Applications that can leverage ML as an innovative differentiator can add a tremendous amount of value to the business, well beyond the investment in ML-enabled systems.

MLops is becoming the hottest career path of late due to the new reliance on AI/ML augmented business systems that drive intelligent supply chains, detect fraud, and provide marketing and sales analytics. Of course, we only have to look at the excitement around ChatGPT to see the interest and potential in weaponizing AI to drive bigger profits, but this has really been evolving for 20 years.

What are the main tasks related to MLops? What would you be working on day to day if you switched to a job related to MLops?

• Model Implementation: deploy machine learning models in a production environment, making them accessible to business applications
• Model Tracking: evaluate the performance of the model once deployed to ensure that it delivers the desired results
• Versioned: keep track of different versions of models as they evolve and improve over time
• Model recycling: update the model with new data to ensure it remains accurate and relevant when data becomes outdated, performance declines, or exhibits bias
• Evidence: ensure that a model works optimally
• Automation: automate tasks such as model deployment, monitoring, and training to reduce the time and effort required to manage models and free up valuable resources for other tasks

Having done each of these tasks at some point in my career, nothing I’ve listed is that hard to understand. MLops is typically part of the existing cloudps team, but will require special training in machine learning in general, as well as company-specific ML systems. Then it’s just a matter of following the processes and procedures to keep the ML system up and running.

Another reason this is becoming a hot job ticket right now: If machine learning systems are not properly operated and maintained, the company can experience major problems. These can range from a misdirected marketing campaign that loses millions of dollars, to lawsuits stemming from bias in a machine learning system that approves or denies home mortgage loans to families. Many things can and will go wrong. Having the right MLops talent will reduce the risk.

Is MLops right for you? If you’re looking for a higher-paying career that requires ongoing new training, and you’re interested in ML as a technology, this might be the most fun and lucrative job you can get right now.