Data Platform Migration for a Healthcare Logistics Company

The situation

A healthcare logistics company coordinates in-home services for elderly and disabled individuals: transportation, grocery delivery, and nursing visits. The company works with insurance carriers and government programs, which means every ride, every service, and every invoice carries reporting obligations tied to plan eligibility, benefits utilization, and regulatory compliance.

The company had built its initial data infrastructure on Snowflake with a set of dbt models handling transformations for reporting. The architecture worked, but the economics did not. Query patterns were periodic rather than continuous, and the Snowflake credit model charged for warehouse uptime regardless of actual utilization. The company needed its data infrastructure costs to scale with its actual usage, not with always-on compute.

At the same time, the internal team lacked the data engineering capacity to manage the migration, maintain the transformation layer, and build the new reports that insurance partners and operations teams were requesting.

What we built

CorrDyn executed the warehouse migration from Snowflake to MotherDuck in 8 days. The existing dbt project provided a clean separation between transformation logic and warehouse-specific configuration, which made the migration tractable on that timeline. We updated connection profiles, adjusted SQL dialect differences, validated output parity between environments, and cut over the production pipeline.

With the warehouse stable on MotherDuck, we turned to the reporting layer. The company insurance partners require regular reporting on member eligibility, benefits exhaustion, and plan utilization. We built dbt models for the core entities the reporting touches: members, assessments, plans, and claims. Shared definitions across reports mean that eligibility numbers in one report match the numbers in the invoicing template, eliminating a class of reconciliation errors that had consumed staff time.

Estuary handles change data capture from the company application database into MotherDuck, keeping the analytical layer current without manual extracts. The pipeline runs on a schedule tuned to balance freshness against cost.

What changed

The migration cut warehouse costs to a level sustainable for the company’s actual query volume. The 8-day timeline meant minimal disruption to reporting workflows during the transition.

Reporting that had required manual coordination between operations, finance, and carrier-facing teams now runs from governed dbt models with consistent definitions. When a new plan requires a new benefits report, the dbt layer provides the scaffolding to build it without redefining core entities from scratch.

The engagement continues as a managed data team on a monthly budget. CorrDyn handles new reporting requirements, pipeline maintenance, and data model development as the company grows its carrier relationships and service offerings.