Application Overview
The C3 Generative AI Application allows you to interact with internal documents and data sources using natural language. It retrieves relevant content, maintains context, reasons over information, and retries queries when needed.
The application simplifies how you retrieve and apply internal knowledge, making it easier to support ongoing operations and informed decision-making.
You can use the application on its own or embed it within other C3 AI applications, such as C3 AI Reliability or C3 AI Demand Forecasting. This guide covers how to use the C3 Generative AI Application as a standalone product.
The following image provides an example of the C3 Generative AI application UI.
Features and capabilities
- Natural language search and chat interfaces - Provides a way for users to interact with the application using conversational language to find information and perform tasks.
- Fully configurable retrieval system - Provides flexibility to tailor the information retrieval process to specific domains, ensuring relevant and precise results.
- LLM integration and hosting services - Supports integration with large language models (LLMs) hosted externally or within the C3 AI Platform to enable advanced text generation and semantic understanding.
- Document and embeddings management experience - Simplifies handling of unstructured documents and vector representations to support efficient storage, retrieval, and analysis.
- Database integrations - Enable connectivity to structured data sources, allowing reasoning across structured and unstructured data.
- Prompt engineering experience - Supports crafting of effective prompts that guide the AI to generate contextually relevant outputs for specific domains or business contexts.
- SDK for full extensibility and augmentation - Provides tools for developers to extend and customize application functionality.
How it works
The C3 Generative AI Application connects to internal data sources such as documents, blob storage, and databases. When a user submits a query, the application performs the following steps:
- Retrieves relevant context from indexed sources.
- Runs configured agents and tools based on the query.
- Generates a response using a language model.
- Filters the result and returns it through the UI.
For details, see Architecture and Capabilities page of the C3 Generative AI Application.