AI provider credentials are supplied through environment variables.
The repository does not track the local .env file.
Only .env.example is included for configuration documentation.
The production Docker image runs the application using a dedicated non-root user.
This reduces the privileges available to the application process inside the container.
InsightForge AI currently supports:
- CSV
- XLSX
Unsupported file extensions are rejected.
Uploaded files are processed within the application's upload storage.
The API uses FastAPI request validation and application-level validation for supported dataset types.
For internet-facing deployments, operators should additionally provide:
- HTTPS/TLS;
- authentication and authorization appropriate to the deployment;
- rate limiting;
- secure secret management;
- network controls;
- centralized logging;
- monitoring; and
- appropriate data retention policies.
These controls depend on the deployment environment and are not represented as additional product functionality in the v1.0.0 release.
Users should avoid uploading confidential or regulated data unless their deployment environment, access controls, contractual arrangements, and applicable data protection requirements have been appropriately configured.
AI-generated reports should be treated as decision-support output.
Organizations should validate important conclusions against source data and appropriate human expertise before making consequential business decisions.