{"id":77785,"date":"2024-07-31T02:39:37","date_gmt":"2024-07-31T00:39:37","guid":{"rendered":"https:\/\/intellias.com\/?post_type=blog&p=77785"},"modified":"2024-08-16T21:16:28","modified_gmt":"2024-08-16T19:16:28","slug":"mlops-for-productizing-ai-the-lean-approach-to-model-development","status":"publish","type":"blog","link":"https:\/\/intellias.com\/mlops-for-productizing-ai\/","title":{"rendered":"MLOps for Productizing AI: The Lean Approach to Model Development"},"content":{"rendered":"

Different types of AI-based solutions (machine learning, deep learning, and foundation models) are taking automation to the next level, offering both opportunities for cost savings and revenue generation. Meanwhile, AI and its \u201clighter\u201d edition \u2014 machine learning \u2014 are only entering the era of productization. Although many data science teams have successfully run several pilots, far fewer have large-scale applications in production. In 2018, only 15% of enterprises had AI models in production for 5+ years, while the majority (49%) were still exploring use cases according to O\u2019Reilly<\/a>.<\/p>\n

\n
\n

A 2023 survey<\/a> by Insider Intelligence found that 42% of North American and 38% of EMEA companies still hadn\u2019t adopted AI or ML or were in the research phase, while 21% and 16% (respectively) were scaling up or already had mature products.<\/p>\n <\/div> \n <\/div>\n

In this post, we look at what\u2019s holding leaders back from productizing AI and ML applications and how MLOps helps bring more pilots to production.<\/p>\n

Why Productizing AI and Machine Learning Models Is Hard<\/h2>\n

Corporate tech budgets have been growing year over year. Gartner<\/a> expects global IT spending to increase by another 8% in 2024 and reach $5.1 trillion. The hype around GenAI has also prompted some 45% of executive leaders<\/a> to further increase their AI budgets. In 2023, over 60% of business leaders named applied ML<\/a> among their top three priorities for the next year.<\/p>\n

Budgets? Check. Strong buy-in? Present. So why aren\u2019t companies progressing as much as they supposedly should with applied AI\/ML?<\/p>\n

\n
\n

Because of its iterative nature, the process of launching AI\/ML applications differs a lot from the process of launching any other type of software.<\/p>\n <\/div> \n <\/div>\n

Standard apps only include code components, while ML models are essentially a bundle of code and data. This adds new layers of complexity:<\/p>\n

Data constraints<\/h3>\n

To produce accurate outputs, ML models require training data that is labeled, consistent, bias-free, and representative of the use case. Without a proper underlying data architecture and appropriate data governance, procuring training and validation data is hard. Acquiring accurate training data is the top challenge for 40% of ML teams<\/a>. Even when appropriate data is available, it still can be off-limits due to privacy or compliance requirements. Finally, when some data is cleared for use, it still has to be transformed, labeled, and stored in a secure repository \u2014 a step that can eat up to 80% of data scientists\u2019 time. On the upside, ML models can be created with limited datasets. Three in five<\/a> tech professionals agree that the quality of training data is more important than the quantity of training data for achieving the best outcomes.<\/p>\n

Lack of processes<\/h3>\n

The standard ML model lifecycle includes more steps compared to the application development lifecycle: data collection and processing, model development, model versioning and integration, plus ongoing model monitoring for performance, security, and effective governance.<\/p>\n

Most of these processes are performed in an ad-hoc fashion with data scientists and operational professionals often using different tools and thus missing alignment. Over 40% of tech teams<\/a> admit that their current process for developing ML models cannot be replicated for another pilot or use case. Subsequently, model development takes a tremendous amount of time and resources. Infrastructure costs run high, while deployed models fail to demonstrate ROI in production.<\/p>\n

\n
\n

Redefining Productivity in the Workplace with AI-Powered Digital Employee Assistants.<\/p>\n

\n
<\/div>\n <\/div>\n <\/div>\n Discover more<\/span>\n\t\t <\/a><\/div>\n

Lack of predictability<\/h3>\n

An ad hoc approach to ML model development creates further issues down the line. Models crash and burn before entering production due to poor (or non-existent) testing. According to various sources<\/a>, data scientists shelve over 80% of ML models before even bringing them to production; only 11% of models are successfully deployed to production. And for those that are deployed, new issues emerge post-deployment. Model drift is inevitable. Explainability, auditability, security, and good governance practices become mandatory, and end-users may require tight SLAs.<\/p>\n

That said, we\u2019re still seeing successfully productized ML and AI models. So what do successful teams do differently? They rely on building a strong machine learning operations (MLOps) process.<\/p>\n

What is MLOps?<\/h2>\n

MLOps is a set of practices and standardized workflows for creating a repeatable, continuous, and automated process for building and deploying machine learning models. Effectively, MLOps service<\/a> adapts certain DevOps best practices like continuous integration (CI), version control, and pipeline automation to the realities of developing different ML (Machine Learning), DL (Deep Learning), and AI (Artificial Intelligence) solutions.<\/p>\n

\"What<\/p>\n

Source: MLOps.org<\/a><\/em><\/p>\n

The concept of MLOps was originally conceptualized by Google researchers in the now-classic paper \u201cHidden Technical Debt in Machine Learning Systems<\/a>.\u201d<\/em> This paper describes at great length all the machine learning model development issues the team was facing:<\/p>\n