C3 AI Documentation Home

C3 Agentic AI Platform Packages Reference

The C3 Agentic AI Platform ships with packages that provide core functionality for your applications. Many of these packages are automatically available without requiring additional dependencies, while others may need to be explicitly added depending on your application's needs.

This reference organizes packages by functional category to help you quickly find the capabilities you need for your application development.

Model Context Protocol (MCP)

These packages provide foundational infrastructure and domain-specific data models.

Package NameDescriptionWhat You Can Build
mcpServerMCP (Model Context Protocol) platform with central server, tool framework, and web-based management UI.Standardized interfaces between AI models and external tools, reusable tool libraries, and managed tool permissions using GenaiCore for AI-powered capabilities.

UI Components

These packages enable you to build user interfaces with React components, visualization tools, and comprehensive styling systems.

Package NameDescriptionWhat You Can Build
uiInfrastructureUI infrastructure for C3 Type-based frontends.Type-driven UIs with UiSdlMetadataLoader for generating TSX from JSON metadata, UiSdlState for Redux state management, and UiSdlConfig for centralized configuration.
uiComponentLibraryType definitions for all UI components.Component-based UIs with 100+ Types including UiSdlDataGrid, UiSdlContentFrame for page layout, and UiSdlAction for Redux integration.
uiInfrastructureReactReact-specific UI infrastructure.React UIs with hooks for state management, render condition Types like UiSdlPermissionRenderCondition, and React component wrappers.
uiComponentLibraryReactReact implementations of C3 AI UI Component Library.Production-ready dashboards, forms, navigation menus, and data visualizations using UiSdlDataGrid for tabular data and UiSdlCardList for card-based displays.
cssLibraryC3 CSS Library with standardized styling.Consistent application styling with themes, responsive design utilities, and standardized component styling.
uiGraphComponent / uiGraphComponentReactGraph visualization components.Network graphs, relationship mapping, and interactive graph exploration with node-based visualizations.
uiTimeGraphComponent / uiTimeGraphComponentReactTime-series graph components.Temporal data displays with interactive time navigation, zoom controls, and time range selection.
uiGanttChart / uiGanttChartReactGantt chart components.Project timeline visualizations, task dependencies, and resource allocation views for project management.
uiSdlDiagram / uiSdlReactSDL diagram visualization.Schema visualizations, type relationship diagrams, and interactive schema exploration.
nodeDiagramNode-based diagram builder.Workflow visualizations and process maps with interactive nodes, connections, and hierarchical layouts.

Machine Learning

These packages provide complete ML lifecycle management from training to deployment, evaluation, and monitoring.

Package NameDescriptionWhat You Can Build
modelOpsModel Operations (ModelOps) platform.End-to-end ML lifecycle management with model registration using ModelOpsApplicationState, deployment workflows, monitoring dashboards, and pipeline visualization.
modelInferenceServiceModel serving infrastructure.REST APIs for serving ML models with preprocessing pipes, model chaining, and postprocessing transformations.
modelInferencePipesML inference pipelines with Transformers and vLLM.Custom embedding pipes using MlAtomicPipe or MlDynamicPipe, NLP tasks using TransformersPipe, and high-performance LLM serving with VllmPipe for async operations.
mlEvaluationModel evaluation framework.Systematic model comparison with MlEvaluation.Job for orchestration, MlEvaluator for defining scope, and auto-logging for metric capture.
mlExperimentTrackerML experiment tracking system.Experiment tracking with MlExperiment.Run for hierarchical organization, 5 plot Types for visualization, and MlExperimentLogger for programmatic logging.
mlHpoHyperparameter optimization framework.Automated model tuning with MlHpo.Experiment for organization, MlHpo.RunCoordinator for parallel trials, and experiment tracker integration.
distributedTrainingDistributed training infrastructure.Multi-node training with DistributedTrainingCoordinator, PytorchPipe for PyTorch models, and scaling across multiple GPUs.

GenAI Platform

These packages enable you to build Generative AI applications with LLM integrations, agents, and vector stores.

Package NameDescriptionWhat You Can Build
genaiPlatformGenerative AI Platform with LLM integrations and agent framework.GenAI applications with GenaiCore.Agent for conversational AI, GenaiCore.Tool for function calling, multi-provider LLM support using GenaiCore.Llm, and vector stores for semantic search.
genAiBaseBase Types and interfaces for GenAI.GenAI foundations with Genai.Agent configurations, Genai.Llm interfaces for multiple providers, Genai.Memory for conversation history, and Genai.Project for orchestration.
genAiSearchC3 Generative AI Search Application.Enterprise search UI with GenAiUiApplicationState for search state management, GenAiUiConfig for UI configuration, conversation threading, document management pages, and customizable search experiences built on top of genAiBase.
generativeAiBase infrastructure for GenAI.GPU-accelerated environments with runtime configurations for OpenAI and Jupyter resource profiles for T4 and V100 GPUs.

Integrations

These packages provide integrations with third-party services.

Package NameDescriptionWhat You Can Build
twilioTwilio SMS and voice integration.SMS notifications, voice calls, and multi-channel communication using Twilio APIs.
googleTagManagerGoogle Tag Manager integration.Analytics tracking, marketing tags, and conversion tracking with Google Tag Manager.
aiOpsServiceAI operations service.AI-powered operations automation and intelligent monitoring.

See also

To learn about the three Types of packages available (built-in, available, and external), see Select Packages Overview.

Was this page helpful?