At Nilus, a fintech startup, I led the development of a customer report page that presented sensitive financial data such as transactions, balances, and money flows. The challenge was that generating these reports on the fly through API calls was resource-intensive, leading to slow system performance.

To solve this, I implemented materialized views to pre-compute and store data in the database. This approach significantly reduced report generation time and system load by enabling data to be fetched directly from pre-aggregated views rather than recalculating it in real time.

However, the static nature of materialized views presented another challenge: discrepancies between real-time data and the report. I addressed this by setting up cron jobs for daily updates and leveraging Apache Kafka to trigger real-time updates when specific actions occurred in the app.

Additional challenges included integrating materialized views into the existing Pydantic-based system, which lacked official support, and minimizing database blocking during view updates. To resolve these, I implemented custom integration solutions and optimized update processes to prevent bottlenecks.

This project improved system performance, ensured accurate and up-to-date financial data for users, and showcased my ability to tackle complex technical challenges in a high-pressure environment.