Modernization of automotive recall campaigns
Optimizing global portal databases helping retailers and cities to monitor the recycling of 50 billion used bottles and cans a year in the world
RecallMasters
CUSTOMER SPOTLIGHT
The customer acquisition and retention platform leverages best-in-class vehicle owner data to connect 2nd/3rd/4th generation owners – 3X more customers and prospects than what is visible to even the manufacturers. Repeat servicing is supported through a targeted retention card effort that broadens the reach to lost and lapsed customers before they defect. The turnkey, fully-integrated customer lifecycle management solution creates greater recall awareness and helps automakers protect their brand.
Project Challenge
The performance of all product features was severely impacted due to slow database queries. The root cause was identified as the data layer, served via a self-managed PostgreSQL database hosted on a shared droplet in the DigitalOcean cloud. As a result, nearly every query was taking an excessive amount of time to complete, degrading the overall user experience.
Solution
Given the severity of the performance issues and their impact on day-to-day operations, we adopted a phased approach to the solution—focusing first on quick wins to deliver immediate relief, followed by long-term improvements to ensure scalability, stability, and maintainability.
The short-term priority was to stabilize the system and reduce query latency without waiting for large infrastructure changes. Once the environment was stable, we shifted our focus to a more strategic, long-term architecture capable of supporting future growth and operational efficiency.
Quick Wins
We began by analyzing the existing PostgreSQL setup and found that the database was running on a shared droplet alongside other compute-intensive services like ETL jobs. This led to significant resource contention.
To immediately reduce system pressure, we migrated the database to a dedicated self-managed virtual machine. This separation improved performance and responsiveness. We also performed a full database audit, implemented maintenance best practices (vacuuming, reindexing), and optimized numerous slow queries through indexing and refactoring.
Long-Term Improvements
After stabilizing the system, we migrated the database to Amazon RDS for PostgreSQL to benefit from managed services, better reliability, and automated maintenance. We introduced horizontal scaling by adding read replicas and redirected reporting workloads and scheduled queries to those replicas, easing the load on the primary instance. Monitoring and alerting systems were put in place for proactive issue detection. Finally, we partnered with the development team to optimize existing application features, refactor inefficient data access patterns, and align the system architecture with modern best practices—ensuring better long-term performance and scalability.