How does the AI import builder analyze my data source?
When you set up a new data import in Analytics → Data Import, the platform uses AI to automatically analyze your data source and configure the connection. Here is what happens behind the scenes:
1. You provide source information
Depending on the import method you choose, you provide different inputs:
- API: Your vendor's Swagger/OpenAPI specification (URL or file upload) and base API URL
- File upload: A file with meter readings
- Email: The sender email address and subject filter for automated reports
- Webhook: An example payload that your BMS or IoT gateway sends
2. AI analyzes the structure
The platform sends your source information to a built-in AI analysis service. The AI reads through the specification or sample data and identifies:
- Which fields contain timestamps, meter IDs, and reading values
- What meters exist in the data (with suggested names, types, and units)
- How to transform the source format into the platform's internal format
- A confidence score (0–100%) for each mapping
3. Connection test
For API sources, the system runs a test request against the real endpoint to verify that authentication works and data comes back in the expected format.
4. Automatic configuration
Once analysis succeeds, the AI-generated configuration is forwarded to the data service, which creates a scheduled runner. For APIs this means automatic fetching on your chosen interval (e.g., every 15 minutes or hourly). For other methods the system registers the appropriate listener.
You do not need technical knowledge to use this. The AI handles field mapping, data type detection, and scheduling configuration. You just provide the source and confirm the results.
Where to start: Analytics → Data Import → select your import method → follow the setup steps.