Deploy, monitor, and re-train models at scale
The C3 AI Platform’s unified interface allows data scientists to review, select, and deploy millions of models into production with a single click.
It’s simple to test competing modeling strategies and deploy the highest performing ones, maximizing business outcomes.
Shift your model to align with business needs by using flexible inference modes. Choose from on-demand computing resources that enable high-throughput batch inference or low latency inference. The C3 AI Platform also supports inference triggered by streaming data, events, or a schedule.
Ensure sustained business value with comprehensive model monitoring. Automated real-time alerts for feature, prediction, label or concept drift monitor models without need for configuration. Teams can also customize alerts with no-code tools and review alert details to conduct root cause analysis, quickly remediating any model drift issues.
Manage production models effectively and quickly with configurable automation. Teams can define automated retraining workloads based on new events, time, and model performance.
Seamless model operations are facilitated by a central repository that stores all model versions and associated metadata. Within the registry, users can explore, re-use, and improve models, review metadata for auditability and traceability, and share models across teams.
Ensure maximum business value from your AI applications and models with comprehensive lifecycle management that automatically stores and versions all models in production, guaranteeing traceability. Users can promote, demote, and retire models to optimize value.