C3 AI Documentation Home

Ask Questions Using C3 Generative AI

You can ask questions to get help and information on C3 AI products using the C3 Generative AI feature in C3 AI Studio. The C3 Generative AI feature has two modes: Ask and Agent.

Use the C3 Generative AI feature

To use the C3 Generative AI feature, select the C3 Generative AI button in C3 AI Studio. A chat box displays and offers example questions to start with.

Select Ask to request general information on C3 AI products, concepts, or tasks. Or, select Agent to query information and data specific to your own deployment.

Ask mode

Use the Ask mode to request general information on C3 AI products, concepts, and tasks.

Here are some example questions:

  • "Tell me about the Demand Forecasting application."
  • "How do I show documentation for a specific Type in C3 AI Console?"
  • "How do I start a single-node environment?"

Ask mode provides general information on C3 AI concepts, products, and tasks. For example, the request, "Tell me about the Demand Forecasting application" returns high-level information about the application, its technology, and use cases.

Because Ask mode does not consider your own cluster, environments, applications, and configurations, completing requests in Ask mode is faster than Agent mode.

Agent mode

Use the Agent mode to query information and data specific to your own deployment.

Here are some example questions:

  • "Show me application logs."
  • "What is the <application name> application for?"
  • "How do I monitor CPU utilization for the <environment name> environment?"

Agent mode considers your own cluster, environments, applications, and configurations. Agent mode is aware of the page you are on in C3 AI Studio.

For example, if you are on a specific application page, Agent mode queries data specific to that application when you ask about it. The request, "Show me application logs" returns a summary of logs for the application that you are viewing in C3 AI Studio.

Because Agent mode provides more robust and deployment-specific responses, completing requests in Agent mode takes longer than Ask mode.

Was this page helpful?