Model Development
The C3 AI Application Platform platform provides data science teams with un-compromised flexibility in building and testing machine learning models based on prepackaged templates that use the latest open-source frameworks and C3 AI’s ML Pipeline technology.
- Use the model-driven architecture to manage and scale AI/ML model development
- Compare models from Python, R, Spark with the same API
- Generate better models using Auto ML capabilities
- Train on all data using distributed architecture of the C3 AI Application Platform platform
ML Pipelines
- Leverage library of 30 pre-built ML pipes on the C3 AI Application Platform and choose from a selection of deep learning, natural language processing, forecasting, and tree-based model pipes
- Link composable ML pipes together to solve more complex uses cases
- Wrap your existing models in a custom ML Pipeline to operate at scale
- Use automatic model management including scoring, interpretability, model version, and validation
Auto ML
- Run pre-packaged experiments to automate and scale-out time-intensive processes like parallelized model training and hyperparameter optimization
- Choose the best approach for your hyperparameter optimization experiment, including randomized, grid, or Bayesian optimization
- Parallelize experiments across the auto-scaling cluster or specify resource requirements such as CPU vs. GPU
- Track all experiments in C3 AI ML Studio’s visual experiment tool