Fixing Search and Data Access Parity
Search results and contact visibility could diverge between SQL and Elasticsearch depending on which path a query took — causing agents and admins to see inconsistent results for the same data. The fix required mapping the full permission and filtering rule set across both systems and aligning their behavior.
My Role
Diagnosed the parity issues, mapped the divergent business rules between SQL and Elasticsearch query paths, implemented the fixes, and validated consistent behavior across the affected workflows.
Problem
Segmentation, permissions, and contact visibility could produce different results depending on whether a query was routed through SQL or Elasticsearch. This created support escalations, trust issues, and incorrect access patterns for shared and pond contacts.
Approach
Traced query paths for affected workflows through both SQL and the Elasticsearch index. Identified where filtering logic, permission checks, and visibility rules diverged. Aligned the index behavior to match the SQL source of truth while preserving query performance.
Outcome
Eliminated the class of visibility and segmentation inconsistencies that had been generating support escalations. Improved agent and admin trust in search results and contact access across the CRM.
Stack
Elasticsearch, SQL, PHP, Laravel
Private internal project
This work was built for internal production use, so source code and detailed implementation materials are not public.