MLOps as a Critical, Emerging Role

Once an organization has the ability to quickly operationalize data projects and moves from a handful to hundreds (or thousands) of machine learning models in production, the question of maintenance and management arise.  Enter: MLOps.

“Technology innovation leaders are keen to apply DevOps principles for AI and ML projects, but they often struggle with architecting a solution for automating end-to-end ML pipelines across data preparation, model building, deployment and production due to lack of process and tooling know-how.”

– Gartner, Accelerate Your Machine Learning and Artificial Intelligence Journey Using These DevOps Best Practices, 12 November 2019, Arun Chandrasekaran and Farhan Choudhary

Watch Video

MLOps help ensure that deployed models are well maintained, performing as expected, and not having any adverse effects on the business. This role is crucial in protecting the business from risks due to models that drift over time or that are deployed but unmaintained or unmonitored. At a time when issues like responsibility and bias are at the forefront, MLOps becomes even more important to close the feedback loop between operationalized models and their impact.

Watch Video

Dataiku for MLOps

• Monitoring: Track and visualize drift over time for all models across the organization in one central location and implement automatic data validation policies.
• Scalability: Leverage existing distributed storage and processing infrastructures (SQL, Spark) to deploy and manage containerized services at scale. Plus, scale job execution and ML model training with Docker  and Kubernetes.
• Code & Integration: Build end-to-end solutions and services from scratch with the languages and tools you already know and love (R, Python, Scala, etc.) with Dataiku’s transparent SDK or leading GUI.
• Operationalization: Automate, operationalize, and monitor data pipelines without having to rewrite custom prediction code or re-think existing infrastructures with Dataiku’s dedicated model deployment API.