Data connectors
Feed document generation from your systems. Beyond CSV and JSON, DocsNG connects to SQL databases, REST/OData endpoints and inbound webhooks, and an AI copilot helps map incoming data to your models.
Built-in inputs
- JSON / REST — the first-class API input; POST a template id and a JSON array of documents. See Generating documents.
- CSV — bulk import via the GUI, with relationship mapping. See Generating documents.
External connectors
| Connector | What it does |
|---|---|
| SQL | Read-only, parameterised queries against PostgreSQL, SQL Server, MySQL or Oracle, mapped to a data model. |
| REST / OData | Fetch records from an HTTP/OData endpoint with authentication and pagination. |
| Webhook | An inbound webhook maps a posted payload to a model + template and triggers generation. |
Each connector maps source fields to a DocsNG data model. Configure a connector, map the fields, run a test fetch to preview, then use it as a data source for generation.
Safe by design: connector credentials are held in the secret store (never in committed config), and target hosts are allow-listed to prevent requests to internal/metadata addresses (SSRF protection).
Data-mapping copilot
When you connect a new source or paste a sample JSON schema, the AI copilot proposes field-to-field mappings to your data model — with confidence scores and type-compatibility notes — and can offer to create missing model fields. You confirm the mapping before it is applied, and it is saved for reuse.
The copilot is optional and air-gap-aware; only schema/sample metadata is sent to the model, never bulk sensitive data.
Need help? Contact us or try the live demo.