# Location Services: Synapse and SQL Server onto Databricks

> Location Services' repossession platform automates assignment intake, so fewer people handle each case, and our engineers built it with the company's own technology team. When it was time to leave Synapse and SQL Server, we moved it onto Databricks in waves that could each be rolled back, and the rules engine that checks every assignment came through unchanged.

Client: Location Services
Sector: Automotive / Fintech
Stack: Databricks Lakehouse, Lakebase, Unity Catalog, Unity Gateway, TechFabric Airlift
Canonical: https://www.techfabric.com/case-studies/vrm

---

## Challenge

**Automate assignment intake, and run the operation with a smaller FTE count**

Location Services needed a new, integrated platform for its asset recovery operations, VRM Pro, to replace older infrastructure and systems.

The project's KPIs were to automate the process flows and reduce the FTE count. The key to both was automating assignment intake, using microservices and newer technology in place of the monolithic platform.

## Strategy

**Design thinking first, and an automated intake engine out of it**

Location Services' product owners and project stakeholders worked through the process with design thinking, and identified the need for an automated intake engine. These were the areas it had to cover:

- Data validation
- Debtor of record
- Person search
- Deceased
- Bankruptcy
- Coverage
- Asset
- Lot check
- Address
- VIN decode
- Fee check

Enterprise architects worked with Location Services' internal technology team to plan the move to the cloud, which changed how the organisation processes repossession orders and transport fulfilment.

## Solution

**A new rules engine to automate assignment intake**

The rules engine automates assignment intake across those areas.

Azure Cosmos DB addressed the data layer. Reporting for VRM Pro was built on Power BI in three parts.

- **Power BI Query** turns the platform's records into tables, links them, calculates the metrics and aggregates them.
- **Power BI Desktop** is where the reports themselves are built.
- **Power BI Service** refreshes each client's reports on a schedule.

## Results

**Early-stage results after the first release**

Shortly after the VRM Pro minimum viable product was released, Location Services saw immediate results and substantial efficiency gains, and the Azure data platform proved to be the difference. New features and improvements since have kept bringing value to the operations team.

- **+$890k** overall financial impact on the business
- **73%** increase in weekly assignments after 30 days
- **47%** increase in weekly recoveries after 30 days
- **51%** increase in weekly recoveries by VIN after 30 days

> "Your team's digital transformation efforts have completely remade our vehicle repossession business. We've experienced unprecedented efficiency gains, reduced operational costs, and a significant boost in recovery rates. It's like our business has been reborn in the digital age, and we couldn't be happier with the results."

> **Client product owner**, vrmPro

The platform was built to scale with multiple integrations, and the business has grown its volume by 250% in three years.

## Since then: onto Databricks

**Synapse and SQL Server onto Databricks, in waves that could each be reversed**

Location Services has moved off Synapse and SQL Server onto Databricks. The warehouse and ETL conversion ran on TechFabric Airlift, which composes Databricks Lakebridge and wraps each wave in signed certificates and a reversible cutover, so the estate could move without a weekend that had no way back.

The operations described above now run on the lakehouse. Lakebase carries the operational data the assignment engine reads and writes, Unity Catalog holds the grants, and Unity Gateway sits in front of the model calls.

The rules engine did not change shape in the move, which is the point: a migration that also rewrites the business logic gives you no way to tell which of the two broke something.