Model Deployment
The C3 AI Platform provides data scientists and developers with an end-to-end ML Ops workflow to rapidly develop, test, and operationalize AI / ML algorithms across the enterprise. Data scientists and developers can promote trained models to production with ease, scale-out up to millions of ML model deployments, test competing modeling strategies, while leveraging continuous analytics processing, auto-training, and model performance monitoring powered by the C3 AI Platform.
Multi-model Deployment
- Promote and track ML Models to production without the technical overhead required by traditional systems. Trained models are immediately exposed through addressable API endpoints
- Scale-out up to millions of ML models within a single application through pre-packaged automation including automatic scoring, compute resource auto-scaling, version management, alerting, hyperparameter optimization, and auto retraining
- Test competing modeling strategies in production through champion-challenger deployments, randomized A/B tests, or shadow models
- Meet all rigorous security and governance requirements with full transparency, versioning, and access control for all modeling artifacts (incl. notebooks, input data, features, models, and outputs)