From source to dashboard — with schema safety at every hop
A walk through how Queryvine fits into a real analytics stack — and where schema drift gets caught before it causes damage. Every step runs before data moves, not after it lands.
What happens when a schema changes
Source connection and initial fingerprint
When you add a source connector, Queryvine immediately fingerprints the current schema — field names, types, nullability constraints, and structural nesting. This baseline is stored as version 1.0 in the schema history database.
$ queryvine connect --source salesforce --name crm_accounts
✓ Connected to Salesforce API v58.0
✓ Schema fingerprinted: 43 fields, 3 nested objects
✓ Baseline stored as schema v1.0Continuous polling on configurable cadence
Queryvine polls each source at the interval you configure — every minute for streaming sources, every hour for stable batch sources. On each poll, it re-fingerprints the schema and compares against the last stored version. If anything changed, the drift pipeline kicks in.
Drift detected — rules engine evaluates
When a schema change is detected, Queryvine evaluates your drift_rules.yaml file in order. Each rule matches on the type of change (field renamed, dropped, type changed) and specifies an action: remap, pause, skip, or pass through. First matching rule wins.
Adaptive action executed before data moves
The action runs at ingest time, before any record reaches your warehouse. A remap translates field names in-flight. A pause halts the pipeline, sends an alert, and waits for your team to confirm the correct action. A skip filters out records with the changed field entirely.
Downstream receives consistent schema
Your warehouse tables, dbt models, and BI dashboards receive data with the schema they expect — regardless of what changed upstream. The schema history is fully auditable: you can see exactly when each change occurred and what action was taken.
Ready to see it in your stack?
Connect your first source in under 10 minutes. Free tier, no credit card.