Cutting a data job's run time by 90–95% for an insurance platform
- reduction in execution time
- 90–95%
Context
At NeoSoft Technologies I worked on Insurance Tech products built as microservices.
From Node.js script to MongoDB shell script
A Node.js script that processed data was performing badly. I rewrote it as a MongoDB shell script so the work ran inside the database, not in application code. Execution time dropped by 90–95%.
Third-party integrations
I integrated and maintained complex third-party APIs across several Spring Boot (Java) microservices. Each integration has:
- Error handling, so a partner failure does not crash our service.
- Retries for transient failures.
- Data validation, so bad partner data stops at the edge of our system.