CML
Continuous Machine Learning
Accelerating Data Science at scale
WHAT IS CML?
Continuous Machine Learning (CML) is a set of tools and practices that brings widely used CI/CD process to a Machine Learning workflow.
Using a validated approach to increasing the agility of software development CML helps organizations to integrate MLOps pipelines on top of the their existing technology stack.
_new.png)
CI/CD
for data science
Use GitLab to manage the end-to-end Machine Learning Lifecycle combined with DVC to enable full data, models and code version control and reproducibility
Flexible Deployment Options
Our CML solution can be easily integrated into your existing technology stack with both
On-premise or Cloud-based deployment options
Automated reporting
Auto-generate reports with metrics and plots in each Git Pull Request. Rigorous engineering practices help your team make informed, data-driven decisions
The AI Academy for you
We can enable programmatic access to your Machine Learning “back-end” by creating automated pipelines (“functions”) that simplify model experimentation and deployment while introducing continuous model performance and retraining
Automation Pipelines
01
MODEL
TRAIN
This pipeline allows to automatically create models. Data scientists needs to do the versioning of the data or the experiment parameters
02
MODEL EVALUATE
This pipeline adds the ability to automatically diagnose and evaluate current models against new data
03
MODEL RETRAIN
This pipeline adds the ability to automatically develop new models should the system detect model drift (lower performance of deployed model on new data)
04
MODEL DEPLOY
This pipeline adds the ability to automatic deploy the champion model into the production environment making the process fully automated and achieving Continuous Machine Learning objective