C3 AI Suite

Machine Learning Services

C3 AI Suite’s ML/AI services are designed for data science teams requiring uncompromised flexibility for experimentation, and unrestrained power, monitoring, and automation for ML operationalization. These services provide a full view into the organization’s raw data sources and transformations using visual tools or a notebook. They can address high-complexity use cases using fully configurable and extendable templates for deep learning, natural language processing, CPU/GPU hardware selection, model auto-tuning, live production A/B testing, streaming prediction services, model life cycle management, multiple language support, composable ML Pipelines, continuous logging, CI/CD, and optimization.

C3.ai Webinar Series

Accelerate Data Science and Scale AI

Model 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.
Feature Engineering
  • Declare machine learning features as composable expressions that are database-optimized and require one-tenth the code.
  • Use ability of features to operate on time-series data, unstructured or text-based data, and structured data.
  • Contribute back to the organization by sharing and extending ML features, minimizing rework and technical debt.
  • Enable comprehensive CI/CD testing support to ensure production models receive high-quality, stable streams of data after every upgrade.

On-Demand Jupyter Service and SDK Connectivity
  • Use on-demand Jupyter environments running in containers that can be flexibly configured with open-source libraries from Conda, PIP, or CRAN.
  • Develop and run custom Python functions that are checked into your GitHub repository, tested using PyTest and CI/CD processes, then executed natively in the notebook or converted to an application API.
  • Bring your own client in situations with remote connectivity, and connect to all C3 AI Suite data and services using Python and R SDKs.
Open Platform
  • Spend more time building better models when using state-of-the-art AI/ML technologies like Spark MLlib, TensorFlow, scikit-learn, cuDNN, SciPy, Caffe, PyTorch, AWS ML, Lex, Polly, Rekognition, Azure ML, H20.ai, Stanford Core NLP, and NLTK in one collaborative data science workspace built for the enterprise.

Open Source Platform