Model-Driven Architecture 
Model Ops
The C3 AI Suite’s continuous analytics processing framework automatically and efficiently updates analytics and machine learning predictions as soon as new data arrives and ensures that end users are always using the latest analytics results. The C3 AI Suite enables developers to quickly and easily configure analytics that will trigger only when relevant new data arrives.
Continuous Analytic Processing
- Leverage distributed stream processing to capture new data and update analytics and machine learning predictions
- Configure specific analytics and frequency to update on arrival of new data
- Optimize computation needs by updating only when sufficient new data arrives
Automate Machine Learning
- Retrain algorithms when needed
- Manage retraining pipelines based on incoming data
- Detect anomalies as they occur when using unsupervised algorithms
- Update the predictions for business users when it is required, using latest data
Model Lifecycle Management
- Build and test machine learning models using prepackaged templates based on the latest open-source frameworks and C3 AI’s ML Pipeline technology
- Version all critical assets, including datasets, runtime, and library configurations, machine learning models, and auto hyperparameter tuning experiments to ensure full reproducibility, progress tracking, ML governance, and compliance
- Establish a regular cadence of production deployments and automate prediction and retraining workloads to ensure the best models are selected and kept up to date
