Structured Data Ingestion
After you deploy the C3 Generative AI Application, select Data Fusion in the application sidebar to connect and manage structured data sources.
The application supports:
- Structured data: Tables from databases such as Snowflake, BigQuery, SAP HANA, PostgreSQL, or files in CSV format.
- Unstructured data: Documents such as PDFs, DOCX files, plain text, or scanned images.
This topic explains how structured data fits into the application and where to configure sources in Data Fusion.
Connect structured data sources
Data Fusion is in Beta. Please contact your C3 AI representative to enable this feature.
Structured data sources are configured in the Data Sources tab of the Data Fusion module in C3 AI Studio. You can complete the following tasks:
Connect an external database such as Snowflake, BigQuery, PostgreSQL, SAP HANA, or ServiceNow. You provide connection details, test the connection, and select tables to ingest. See Select and Modify Data Sources for Data Integration.
Upload a CSV file if you prefer not to connect a live database. You select files, define schema fields, and preview or update table metadata. See File Upload Configuration Example.
After you add or upload a data source, select Sync Changes to activate it for queries in the Generative AI Application.
When you configure sources in Data Fusion, you can browse and preview data, add new sources, define schemas and aliases, manage connectors, and troubleshoot issues. See the above linked topics for detailed steps.
Finalize your structured data setup
After syncing your data sources, return to your application and select the Sync Data Model banner to apply the changes. This sync makes all new sources available, and the structured query tool in the Generative AI Application can then retrieve tabular data. You will be able to query your connected databases once the sync completes.