Adding a new data source is three steps: configure the connection (API or CSV), map source fields to canonical schema, set the conflict-resolution priority. Golden DB starts ingesting within minutes.
Step 1 — Configure the connection
Data → Sources → New Source. Pick an API integration (for live providers like ZoomInfo, Apollo, Cognism), a CSV upload (for one-time imports), or a webhook (for custom systems pushing data to turgo).
Step 2 — Map source fields to canonical schema
The mapping screen shows source fields on the left, canonical schema on the right. Drag to map. Use built-in transformations where formats differ (e.g. source phone is "+1 (555) 123-4567" → canonical is "+15551234567").
Step 3 — Set source quality and priority
Per field, set the source's quality score (0–100). High = trust this source for this field; low = use only as fallback. Quality scores drive conflict resolution when this source disagrees with others.
Step 4 — Test with sample data
Click Test → Run sample. The connector pulls 10 records, applies the mapping, and shows what Golden Records would look like. Review for correctness and fix mapping issues before going live.
Step 5 — Enable and monitor
Enable the source. The initial backfill runs (can take hours for large sources). Monitor Data → Sources → [Source] → Ingestion Log for any errors or unmapped fields.
