This section addresses frequently asked questions when connecting data sources, setting up AI/ML model endpoints, or troubleshooting integration issues within AI Squared.

​
Data Source Integration

​
Why is my data source connection failing?

  • Verify that the connection credentials (e.g., host, port, username, password) are correct.
  • Ensure that the network/firewall rules allow connections to AI Squared’s IPs (for on-prem data).
  • Check if the database is online and reachable.

​
What formats are supported for ingesting data?

  • AI Squared supports connections to major databases like Snowflake, BigQuery, PostgreSQL, Oracle, and more.
  • Files such as CSV, Excel, and JSON can be ingested via SFTP or cloud storage (e.g., S3).

​
AI/ML Model Integration

​
How do I connect my hosted model?

  • Use the Add AI/ML Source guide to define your model endpoint.
  • Provide input/output schema details so the platform can handle data mapping effectively.

​
What types of model endpoints are supported?

  • REST-based hosted models with JSON payloads
  • Hosted services like AWS SageMaker, Vertex AI, and custom HTTP endpoints

​
Sync & Schema Issues

​
Why is my sync failing?

  • Confirm that your data model and sync mapping are valid
  • Check that input types in your model schema match your data source fields
  • Review logs for any missing fields or payload mismatches

​
How can I test if my connection is working?

  • Use the “Test Connection” button when setting up a new source or sync.
  • If testing fails, examine error messages and retry with updated configs.