Automated data quality enforcement
Enforce data consistency automatically through matching rules that detect duplicates, stage flagged records for review, and block unvalidated data from reaching production.
Overview
Duplicates and inconsistent records degrade every downstream process: scoring models produce unreliable results, work orders target the wrong assets, and reporting loses accuracy. PTC Orbit addresses this with an automated data quality pipeline in Data Foundry that intercepts incoming records, evaluates them against matching rules, and flags potential duplicates before they enter production.
The pipeline operates continuously. Every record that enters PTC Orbit through data import, sync pipeline, or manual entry passes through this quality gate. Administrators configure the rules. The platform enforces them without requiring manual comparison of each record.
How it works
Data quality enforcement follows a three-stage process.
1. Rule configuration. Administrators define matching rules in Data Foundry that specify which fields to compare and what logic to apply. For example, an asset matching rule might compare serial numbers across incoming and existing records to detect duplicates. PTC Orbit also ships with pre-configured matching rules that activate automatically when Staging Review is enabled for the Asset or Product object.
2. Automated evaluation. When records enter the staging environment, active matching rules evaluate each record against the production database. Records that trigger a match are flagged and routed to the Matching Data Review page. Records that pass all rules proceed to production without delay.
3. Manual review and resolution. Flagged records appear on the Matching Data Review page for human inspection. Reviewers evaluate each flagged pair and decide whether to merge, keep both, or discard the duplicate. After resolution, validated records sync to production.
Pre-configured rules for Asset and Product objects activate through an event handler when Staging Review is toggled on. These rules start in an inactive state: an administrator must activate them explicitly after reviewing the rule configuration.
Key capabilities
• Rule-based duplicate detection. Compare fields such as serial numbers, product codes, or account identifiers across incoming and existing records. Each rule specifies the target object, comparison fields, and matching logic.
• Pre-configured rules. Built-in matching rules for Asset (serial number) and Product (product code) objects eliminate initial setup effort for common use cases.
• Staging review gate. Flagged records are held in staging and do not reach production until a reviewer approves them. This prevents duplicates from propagating into operational data.
• Rule lifecycle management. Activate, deactivate, update, or delete matching rules as data sources evolve and quality requirements change.
Considerations
• Matching rules evaluate records during import. High-volume batch imports with strict matching criteria can increase processing time. Schedule large imports during off-peak hours.
• Disabling Staging Review does not delete pre-configured matching rules: it deactivates them. Re-enabling Staging Review reactivates the rules.
• Custom matching rules require careful field selection. Overly broad criteria flag too many false positives. Overly narrow criteria miss genuine duplicates.
Related Topics