Data health

How the platform validates your data before and after connection, and how to monitor freshness, quality, and sync status.

The platform validates your data automatically and keeps watching it after you connect. Health lives in one place, Settings → Data Health, and answers three questions: is the data fresh, is it complete, and are the syncs running.

Validation happens twice

Before connection

When you add a source, the platform validates the schema: required tables exist, required columns are present, types are compatible, and a sample reads cleanly.

Continuously after

Every sync re-checks freshness, completeness, format compliance, and referential integrity, and flags what it finds.

What the rules check

Check	Requirement	Severity
IDs present and unique (accounts, contacts, events)	100%	Critical
Email present on contacts	above 90%	Critical
Domain present on accounts	above 80%	Warning
Email and domain format valid	pattern match	Warning
Timestamps present, none in the future	100% present	Critical / Warning
Contact-to-account links valid	foreign keys resolve	Warning

You can add custom rules (ranges, patterns, foreign keys, or any boolean SQL expression) per source under Settings → Data Health.

Freshness

Each table carries a warning and a critical threshold. Sensible defaults:

Table	Warning	Critical
Accounts, Contacts	1 hour	4 hours
Events	15 minutes	1 hour

Stale data usually means the upstream warehouse job is delayed, not that the sync broke. Check your warehouse scheduler first.

Sync status and failures

The dashboard lists every inbound and outbound sync with its last run, duration, and row count. When a sync fails, the platform shows the error, the records affected, and the auto-retry schedule; you can retry immediately, view logs, or pause the sync.

Common causes and fixes:

Issue	Cause	Fix
Stale data	Warehouse job delayed	Check the warehouse scheduler
Quality drop	A bad import	Review recent imports
Sync failure	Expired auth	Re-authenticate the connector
Validation errors	Schema change	Update the mappings

Alerts

Health alerts go where your team works: Slack channels or email, with immediate notification on critical failures and a digest for warnings. A daily health report summarizes freshness, quality, and sync status. Delivery preferences are configured in Settings.

Outside the app

Health status can be read outside the app via API and MCP as well.