2 years ago

Category: Software, Data, Microsoft, github, gitlab, automation, bitbucket, artificial-intelligence

Machine learning operations (MLOps) in the last year has emerged as a distinct IT discipline for building machine learning (ML) or artificial intelligence (AI) models. While at first blush that may seem like a viable method for automating the building of AI models, in reality purveyors of MLOps platforms have a vested interest in convincing organizations to acquire platforms that exist outside of best DevOps practices that have already been proven to accelerate application development.

That unsatisfactory experience led to the launch of opensource Data Version Control (DVC) and Continuous Machine Learning (CML) tools that integrate ML workflows into best practices for software development.

They integrate ML workflows into current practices for software development in a way that eliminates the need for many features of proprietary AI platforms such as AWS SageMaker, Microsoft Azure ML and Google Vertex AI by extending traditional software tools like Git and CI/CD platforms to meet the needs of ML researchers and ML engineers.

We at Iterative.ai extend existing DevOps tools so that developers can version data with their code in addition to ML models.