Skip to main content

​
Merge

Create golden records by intelligently combining matched records into a single, authoritative view. The Merge module applies survivorship rules to determine which values become the master record, producing the trusted data your organization needs.

​
Why Merging Matters

After matching identifies related records, you need to combine them into a single, reliable golden record. But which values should survive?
  • Should the most recent address win, or the most complete?
  • What if one system has the SSN and another has the email?
  • How do you handle conflicting values for the same field?
skyMDM’s Merge module answers these questions with configurable survivorship rules that produce consistent, trustworthy master records.

​
What You Can Do

Define Survivorship Rules

Configure which source values survive the merge based on recency, completeness, source priority, or custom logic.

Review Before You Commit

Open a merge candidate to see the suggested golden record beside every source record that fed it, before anything is written.

Handle Conflicts

Fields that disagree across a cluster are highlighted in the review view, and the alternative values are recorded on the golden row for later audit.

Maintain Lineage

Track which source records contributed to each golden record for full traceability.

​
Survivorship Strategies

Survivorship is configured per semantic attribute on the Merge rules page. Any label mapped from two or more tables appears there and can be given its own strategy.

Not Prioritized

Take the first non-null value encountered — the default where no source is more trusted than another

Source Priority

Order the contributing tables explicitly, so the highest-priority source’s value wins

Most Recent

The value from the source record with the latest update timestamp wins

​
Key Capabilities

​
Field-Level Survivorship

Configure different survivorship strategies for different fields. Use recency for addresses, source priority for clinical data, and completeness for demographics.

​
Golden Record Creation

Produce a single master record for each matched entity with the best available values from all contributing sources. Maintain bidirectional links between golden records and their source records. Always know where data came from.

​
The Review Queue

Clusters below the auto-merge threshold pause as merge candidates instead of being written. A steward opens one and sees every source record side by side with differing fields highlighted, the confidence score and the band it fell into, a plain-language summary of why the records matched, and the golden record skyMDM would produce. From there they can approve it as-is, reject it so each source record stands alone, split it to keep only some records together, or divide it into multiple groups when the cluster turned out to hold more than one person. A previous approval can be undone, a comment is required on every decision for the audit trail, and up to fifty candidates can be approved in a single batch.

​
Golden Record Provenance

Every golden record carries how it was created — auto-merge, steward approval, rejection, or split — along with who approved it and which merge candidate produced it, so any record can be traced back to the decision behind it.

​
Consistency Safeguards

The golden record is written to your warehouse first, and only then is the approval recorded. A background reconciler compares recorded decisions against the golden table and heals any drift, so a network failure mid-approval never leaves a half-approved candidate. Merge also refuses to start while Match is still running on the same profile, so golden records are never built from a half-finished cluster set.

​
Real-Time 360° Views

Access the complete, merged view of any entity from the Profiles page, where golden records are searchable and available to downstream systems.

​
Runs Where Your Data Lives

Merge executes natively on Databricks, Snowflake, or Microsoft Fabric, writing golden records into your own catalog or lakehouse. Review Queue approvals write through the same path, so stewarded decisions land in the same golden table as automatic merges.

​
Business Impact

Single Source of Truth

One authoritative record for each patient, provider, or member.

Data-Driven Decisions

Trustworthy data enables confident business and clinical decisions.

Operational Efficiency

Eliminate time spent reconciling conflicting data across systems.

​
Who Benefits

  • Clinical Teams: Access complete, accurate patient records for care delivery
  • Executive Leadership: Trust the data behind dashboards and reports
  • IT Teams: Provide a single API endpoint for master data across the organization
  • Compliance Teams: Demonstrate data accuracy and lineage for audits
  • Analytics Teams: Build reliable models and reports on trusted golden records

​
The MDM Journey Complete

Merge represents the culmination of the skyMDM pipeline:
  1. Connect - Bring data together from disparate sources
  2. Clean - Standardize and correct data quality issues
  3. Validate - Ensure data meets quality thresholds
  4. Enrich - Add context and additional attributes
  5. Map - Apply semantic labels for consistency
  6. Match - Identify related records across sources
  7. Merge - Create authoritative golden records
The result is a unified, trustworthy view of your healthcare data—ready to power better decisions, better care, and better outcomes.